Sarkis Mazmanian, Microbiologist, Medical Investigator, and Biotech Founder

Early in his research career, Sarkis Mazmanian realized that certain bacteria under certain conditions conferred benefits to their hosts. Exploring this beneficial relationship, which is the centerpiece of all Mazmanian's work, has demonstrated profound implications both for fundamental research and in the pursuit of applying these findings to the ultimate goal of improving human health. For Mazmanian, who favors the humble phrase "mouse researcher" as his scientific identity, the foundation of basic science always comes first, and any ideas that can be applied within the context of business ventures and clinical applications flow only from the basis of curiosity-driven science. For all of the enjoyment and sense of accomplishment Mazmanian has felt in launching biotech startups, his true passion is rooted at Caltech, for the mentorship he can provide his students, for the knowledge his students can share with him, and for a laboratory environment where the vastness of the microbiological unknown provides both exhilaration and opportunity.

Beyond the lab, Mazmanian feels a strong connection to his cultural heritage. Born in Lebanon to an ethnic Armenian family, Mazmanian came to the United States as a toddler, where his parents sought a new life and where they could join a growing Armenian community in Los Angeles. In his early years, Mazmanian's experiences were somewhat insular; Armenian was his first language, he attended Armenian schools, and much of his social life revolved around family gatherings. He attended public high school in Los Angeles, and he enrolled at UCLA, where he began his undergraduate studies without a particular interest in science. As Mazmanian remembers it, for one biology class, he found himself turning pages, wanting to see how "the story ended." From there, Mazmanian grabbed onto a curiosity of the microbiological world that has guided his research ever since. Working with Olaf Schneewind in graduate school, Mazmanian learned how to design experiments on staph aureus mutants and manually sequence DNA at a time when automation sequencing techniques were not yet available. For his postdoctoral appointment, Mazmanian worked at Harvard Medical School, where he became fully enmeshed in questions about the microbiome, and why gut bacteria are tolerated and can even be beneficial to their host organisms.

In exploring such fundamental questions with answers that could sometimes feel within reach (and sometimes not), Mazmanian knew that his ideal academic environment would embrace the unknown, would provide the resources necessary to investigate extraordinarily complex problems across a range of biology sub-disciplines, and would nurture opportunities whenever the science suggested possibilities for translational research. In the discussions below, Mazmanian recalls fondly his job talk at Caltech, where the faculty's enthusiasm for his research approach - and in particular, the feeling of adventure they conveyed in sensing that Mazmanian was on to something quite interesting - convinced him that Caltech could be his ideal intellectual and academic home. And staying close to family had its obvious benefits as well. In more recent years, Mazmanian has explored potential connections between gut health and a range of neurological disorders, and his reflections on being less sure over time in those connections offers a powerful lesson in science: pursue the truth by following the data and not preconceptions, and sometimes the value of learning more is in revealing just how much more there is to learn.

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Friday, April 19th, 2022. I'm delighted to be here with Professor Sarkis Mazmanian. Sarkis, great to be with you. Thank you for having me in your office today.

SARKIS MAZMANIAN: Thank you for spending the time with me.

ZIERLER: Sarkis, to start would you please tell me your title and affiliations here at Caltech?

MAZMANIAN: Title is the Luis & Nelly Soux Professor of Microbiology. I'm a professor in Biology and Biological engineering.

ZIERLER: Let's start first with the named chair. Tell me a little bit about the Souxs and if there is any relationship specifically with the research that you do.

MAZMANIAN: I've only met them once and after they endowed the professorship. The Souxs were interested in supporting microbiology research at Caltech and provided the funds for the named professorship at the same time I was just promoted to tenure. The division chair at the time, Steve Mayo, felt that I was worthy, I imagine, of a named professorship and offered that to me of course. I delightfully accepted. I met the Souxs a few years later. They are Venezuelan. Luis is an engineer. He was building mining platforms and other infrastructure for refining in Venezuela, but obviously the social changes in Venezuela were an issue. He subsequently moved to Chile. When I met him, him and his wife were here in Los Angeles on a personal trip from Chile and decided to come by the campus.

ZIERLER: Is there an additional affiliation with the Heritage Medical Research Institute?

MAZMANIAN: There was and for six years, I was an investigator for the Heritage Medical Research Institute. This is an institute that is owned and funded by Richard Merkin who is a trustee here. He's funded a number of different laboratories over the years. It's a program that is still ongoing. I was one of the initial recipients of the investigator award. It's essentially unrestricted funds that allow I think a handful—six to seven—researchers at any one point in time at Caltech to take big risks, take chances in testing hypotheses or concepts that aren't fundable through traditional funding agencies, don't require lots of preliminary data. For the six years that I was receiving HMRI funds—which ended I would say three or four months ago—I think we leveraged that to really expand our research and move it into what I believe to be exciting new directions.

ZIERLER: Is the Heritage Medical Research Institute, is it set up similar to HHMI where it allows funding and opportunities that might not exist otherwise?

MAZMANIAN: It was modeled after HHMI except for the fact that it was originally meant to not provide long term support. It was two to three year appointments, but the concept was very similar to HHMI.

ZIERLER: Sarkis, a question just in light of where we are right now. Your office is in the Chen Institute, of course. Besides the administrative reasonings behind that, what are some of the deeper scientific underpinnings that your research would be in an institute dedicated to neuroscience?

MAZMANIAN: The building is nominally a neuroscience institute. I'm a microbiologist and I think we'll probably get into my training and the trajectory of my career, but we've become quite interested in the gut-brain connection maybe as a general theme for the lab and we have learned how microbes that live in the gut access the brain through a number of different processes and change or alter behavior and affect neurogenerative diseases. Again, I'm not a neuroscientist. I'm learning; my students teach me neuroscience almost on a daily basis. I think our work fits into the broader theme of science within the building and the Chen Institute, but I'm hopeful—I don't know—that the administrators who decided that our lab would move from the Church Labs to Chen—it was not my decision—saw that we provide a unique perspective to neuroscientists.

ZIERLER: The blood-brain barrier as we know it, given what you do and the fact that there are these connections with the brain and the gut, what's problematic or might be a misnomer about the term where there is a barrier that might not exist in actuality?

MAZMANIAN: I mean we know it doesn't exist. It's a barrier because…the blood-brain barrier does keep toxins and other components that are in the bloodstream out of the brain because the brain isn't immune privileged space. It also keeps pathogens out, virus and bacteria. Of course we know nutrients get in and out of the brain, signals get in and out of the brain. It's quite intuitive. What our research has shown is that microbes can access particular portals that they use to send signals through the blood-brain barrier into the brain, as I said, to change the behavior of cells in the brain—neurons and other cells—and impact behaviors.

ZIERLER: There are so many scientific disciplines that you're involved in. There's biochemistry, there's biophysics, there's evolutionary and organismal biology, microbiology, neuroscience. It could go on and on and on.

MAZMANIAN: Immunology and computer science and metabolism that we incorporate in almost every project.

Microbiology as Umbrella Discipline

ZIERLER: What at the end of the day if you had to call yourself a particular kind of scientist either as an umbrella term, either as from your educational trajectory, or your sensibilities—if you only had to pick one what would it be?

MAZMANIAN: If I had to pick one I'd go to my roots as a microbiologist because we view these problems through a microbial lens. Again, applying a broad number of research approaches to our studies…at the end of the day we're generalists. There's almost a method to this madness, right? I started realizing many years ago that education, especially higher education, in Western societies really almost leads us to become quite specialized. I think that's because there is such a broad knowledge base and we have so many tools to work with within each individual discipline that it consumes our education, it consumes our training.

For example, my degrees are in microbiology. Essentially all I learned was microbiology both as an undergrad and as a graduate student. Of course I was exposed to other disciplines, but not formally. And I had to work hard to carve out time to learn even immunology, let alone metabolism by chemistry, among other disciplines. This also happens at the university system—not at Caltech—but at most universities where we silo individuals in one discipline in a building or a floor because we co-house microbiologists all together. What that does—and I've talked to my colleagues at other universities about this—is that it really consumes your bandwidth. If all your interactions are with microbiologists you don't really have time and there's a barrier to go and read or hear about other disciplines. What I realized is that that's not how the human body works and it's really the core of our interest is trying to understand fundamental principles that will help, hopefully someday, with human health and all the organ systems, all these different mediators, molecular mediators, cell types, tissues, what have you—however you want to slice and dice disciplines—are all interacting with each other in real time. One of the things that I started incorporating into our research—again right around the time I started at Caltech—was to make sure that we had essentially a multidisciplinary focus to our work so we can try and piece together all the different reactions going on in the body at least from an intellectual standpoint and hopefully from a practical standpoint.

ZIERLER: The challenge obviously given how multidisciplinary you are is that it's a fire hydrant of information coming at you.

MAZMANIAN: [laugh]

ZIERLER: How do you decide—because resources are limited in terms of personnel, in terms of time—what to segment out in terms of focusing on a paper or a research project?

MAZMANIAN: I think that for me the problem is not so much the fundamental knowledge base, but keeping up with the evolving literature and really trying to know what are the most timely or the most relevant interactions in any of these given disciplines and to what degree we emphasize or deemphasize the focus of those disciplines in a given research program. I have to say to you it's probably case by case because at the end of the day the trainees have a huge voice in the trajectory and the path that they take for their projects. I have a knowledge base. The individual students and trainees have a knowledge base. The lab is eclectic, as I mentioned, in terms of the training background of individuals and then we rely heavily on collaborators which have domain specific knowledge. We take the research or follow the research where it goes without trying to sort of superimpose if this is more of an immunology versus a metabolism project. It's where is the most exciting science? What are the most exciting leads, if you will, to follow? Then try and backfill that concept with information from various different sources.

ZIERLER: As you've alluded, Caltech is unique in the way that these disciplinary boundaries play against the effort to try to make the institute as flat as possible, as multidisciplinary as possible. Beyond your own research group whether here at the Chen Institute or across campus, what are some of the most important collaborations both at the faculty and the research group level for you?

MAZMANIAN: You mean outside of Caltech?

ZIERLER: No. I mean within Caltech, but outside of your research group here.

MAZMANIAN: We collaborate…I would say a third if not half of our projects are collaborative with Viviana Gradinaru who is just two doors away. A neurotechnologist I think is the way she describes herself, but really a brilliant and well-versed neuroscientist. She's sort of my anchor, if you will, on the neuroscience side. I'll take a step back and maybe we'll get to some of the other stuff in subsequent discussions. When I started here at Caltech as a microbiologist studying essentially immunology at the time I really became fascinated with potentially applying some of the concepts we were learning on micro-immune interactions to micro-nervous system interactions based on a really wonderful collaboration with Paul Patterson. I don't know if you know Paul.

ZIERLER: Yeah.

MAZMANIAN: He passed away in 2014. Again maybe I'll tell a fuller story later, but it was really just this in terms of when we first started chatting that led to a collaboration. This awareness that if we bring two disciplines together at the interface of that discipline we can have breakthroughs. We can have breakthroughs not just because of the obvious where we're applying different technologies and different knowledge bases, but what I realized is different disciplines almost speak different languages and approach problems differently. I think it gave us a broader toolkit when we employed perspectives from microbiology and neuroscience. A broad toolkit and a broad perspective with which we can ask questions and potentially answer them.

Getting back to your question. In addition to Viviana, we've collaborated almost episodically with others. Pamela Bjorkman, we've had a number of collaborations with. Mikhail Shapiro, recently we've had collaborations with. Rustem Ismagilov. Again a number of other individuals, but those have been like—and this goes back to a previous comment—when the research takes us into a direction that we're unfamiliar with we reach out to a specialist in that space and of course being at Caltech they welcome us with open arms and they help us solve those problems.

ZIERLER: [laugh] Sarkis, there's so much exciting happening in computation around Caltech right now. How do you slot into that? What are some of the advances in computation and computational biology that are allowing you to pursue questions that might not otherwise be possible?

MAZMANIAN: It's really been in terms of advances in bioinformatics. We're getting into more machine learning and artificial intelligence and being able to predict outcomes based on datasets, but we work in an area where we're still generating data; we're still gathering data. It becomes really hard to impose those predictive models. Really the bioinformatics that allows us to essentially deconvolute large datasets has been an enabling tool that we've used in our research. What do those large datasets look like?

Again, working on gut bacteria, the concept of metagenomics or the ability to broadly sequence tissue and to look at all the different microbes, at least the DNA from all the different microbes, in a sort of complex environment like the gut. I think it's really been transformative to the field because again you can get a fingerprint of who is there. That fingerprint changes. So, David, you and I have different microbiomes. If you were to sequence your sample and mine you'd see something very different, but obviously you and I are more similar to each other than our closest nonhuman primates or even people who live in different geographies and eat different diets.

Those large datasets again are telling, but you have to be able to really mine them at the end of the day and try to understand what is the underlying biology. I can get a little bit more into those details as well. Again, that's something, the metagenomics and the broad sequencing of microorganisms, is something we've been doing for about five or six years now. More recently we've become interested—because the technology has allowed us to—in what are those organisms making? What are the products from those organisms? Because at the end of the day those molecules that the bacteria make are really what's interacting with us and the environment around them.

We've coupled metagenomics to two newer technologies—again on the microbial side—and that's metatranscriptomics and metabolomics. What metatranscriptomics essentially are are the expression of genes within those metagenomes. Again, since metagenomes are DNA, the metatranscriptomics tell us, give us a snapshot, of what is being expressed, which of those genes are being expressed at any one given point in time or in what tissue or in what context. The metabolomics which broadly are small molecules, lipidomics and proteomics as well, give us a snapshot of the molecules, the actual products from those bacteria. Now you can trace the DNA of who's there, the RNA to what's being made, or to what genes are being expressed, and then to the actual products of the molecules being made.

Now again it's not that it's three large datasets, but these datasets are linked to each other because one comes from the other. Really the bioinformatics has allowed us to understand the fingerprint and the biology in our experimental systems, but to also understand how those sequences are linked to each other that allow us to try and understand how the ultimate products of those organisms are interacting with the host. Then we collect a lot of additional datasets on the host side, some of which we also employ bioinformatic tools as well. It's really interesting because each of these different datasets, there are pretty good computational tools to essentially decode the data, but there aren't really good tools to then understand how those datasets are linked to each other. We're not the ones pushing on that frontier. We have some in-house capability, but we have collaborators outside of Caltech who I believe are really experts in trying to piece together all of these different puzzles. Each puzzle is super complex, but piece them all together in a way that makes biological sense.

Basic Science as Foundation to Translation

ZIERLER: Sarkis, an overall question as it relates to your motivations in establishing what problems to work on and really the overall thrust of your research agenda. There's basic science, just figuring out how stuff works and then there's the translational, you see a problem and there's a pathway to a goal toward achieving it. When is it about the basic science and when is it about you have an opportunity really for a translational breakthrough?

MAZMANIAN: It's always about the basic science. That's the foundation. We try and understand, try and take apart these natural systems and try to understand the interactions again between microbes and let's say immune cells or microbes and the nervous system. But we do this in a way, David, that I think we can maybe get more mileage out of our basic discoveries in that we look for these fundamental interactions in models of disease and animal models of disease because there are differences between health and disease in the experimental system. We're exploring this new biological space, and the way we set up these experiments is we explore the space in what are relevant disease models, so I think you sort of get a two-for-one where we can identify novel interactions but do this in a context where you have a disease state as well. Of course, every experiment is controlled where we have "healthy animals" and "disease animals" and so the comparison gives you insight into not just the biological system, but how those biological systems can become dysregulated in the context of disease. Then that allows us to then make predictions of how we can either augment or restrain particular interactions that may potentially give us insights into ameliorating or addressing disease.

ZIERLER: Starting from the framework of basic science, where is that ultimately translated into therapies and to partnerships with hospitals? How does that progress happen from your lab?

MAZMANIAN: It's happened in three different ways. The first is the data are public. They're made available to the world and other people use the data—whether they're academics or people in industry—to either make predictions or to augment their own datasets and move their own projects along. I assume that's happening in the background to some degree. The second is that we've just licensed technology to pharmaceutical companies on a few occasions and the pharmaceutical companies can then develop the process by which then they'll take that information and embark on this long journey—that I'm still learning about frankly—towards generating a therapy, getting the approvals, and the process of actually taking that therapy into humans.

The third is the path that I think most of our research has been translated into are companies that I've started. As you know Caltech's a very entrepreneurial place. I've started three companies since I've been here and my companies have taken technology from our lab. Again, really basic information. Oftentimes nothing more sophisticated than a mouse. Oftentimes less sophisticated in cell culture systems. But taking that information and really asking how do you build on this foundation towards, let's say a molecular therapy? Maybe this is intuitive, but I'll mention it.

Our research has never led to devices or polymers or even diagnostics yet, but it has really led to molecular therapies like drugs at the end of the day. To take basic information, for example these two cells interact, it leads to autoimmunity, so if you prevent these cells from interacting you may be able to prevent or slow down autoimmune reactions. Well, how is that done? What is that drug going to look like? How is that drug going to be manufactured? How is that drug going to be delivered? How is that drug going to be dosed? How are you going to get the regulatory approvals to advance the clinical programs? Where do you get the capital to do so as well? How do you ultimately design a well-powered clinical trial where at the end of the day you're left with solid results as opposed to equivocal outcomes? Then ultimately someday—this has not yet happened—but then marketing that drug. It's an art, it's a science, and it's sort of a long and treacherous road that I'm learning. For example, one of the companies that I've started called Axial Therapeutics now has three drugs in the clinic, two for autism, one for Parkinson's disease. Again, all based on ideas we had 10 years ago and then taking advantage of the technology that we developed in the subsequent years."

ZIERLER: When have you made the decision to start your own company and when have you decided this is public domain, this is something that can be licensed out to other companies?

MAZMANIAN: The public domain part…the dirty, little truth is that Caltech owns everything, right?

ZIERLER: Right. [laugh]

MAZMANIAN: It's not a dirty, little truth, but Caltech is remarkable in the way…and look. Universities just don't make potentially valuable technology public.

ZIERLER: Sure.

MAZMANIAN: They can make datasets public after you've protected them with intellectual property. My decision has been based on two variables. Initially, this is, what are those technologies that I believe would be most valuable to develop? I think objectively speaking not all technology is equal. Some are of potentially higher value than others. Again that's in some way a judgment call based on the context. The other is bandwidth. To be frank with you, I'm invested in my companies in terms of my time and not just my emotions. For example, just today I probably spent two to three hours on my company. I generally reserve Fridays for my consulting work and there's only so much time that I have. I have two young kids and they require a lot of my time. I enjoy spending my time with them. Maybe to trivialize this, my wife told me recently that she will not allow me to start another company. [laugh]

ZIERLER: [laugh] Until the kids are older, maybe.

MAZMANIAN: Or we get an exit for one of them, right?

ZIERLER: It's great to hear you talk about how Caltech is an entrepreneurial place. I'm sure you've heard stories 40 or 50 years ago. People like Lee Hood…that certainly was not the case. It was not an entrepreneurial place. Have you been at Caltech long enough where you've seen that transformation yourself or is your sense that mostly happened by the time you joined the faculty?

MAZMANIAN: It's evolved since I've been here, but I've heard stories. I'm not sure I have a full picture of the Lee Hood years. Maybe you can tell me some of what was going on under the Hood. [laugh]

ZIERLER: [laugh]

MAZMANIAN: There was a pun there, right? Full disclosure, I know Lee pretty well, but when I met him he had already left Caltech and I've heard about some of the tensions that he had here and that the administration had with him. But whether or not that's emblematic of what Caltech was previously, that I cannot speak to. By the time I arrived here—and again maybe I'll tell a personal story. The head of the Office of Technology Transfer was a person named Larry Gilbert. I don't know if you know Larry or not.

ZIERLER: I know of him. Yeah.

MAZMANIAN: He was just a remarkable man. For reasons I don't understand he kind of took me under his wing. There's probably a 50 or 40 year age gap there. He really encouraged me to reach for the stars in terms of the translational potential of our research and let me know that he would do the right thing with that technology. He even one time said that…he was really bullish about what we were doing in autism. Again, initially with Paul Patterson. He said that, "We'll find a good home for this technology. If we can't then we're going to make it publicly available to the world because it's important work." I think he was very pleased that I was able to start a company around that technology and I think he was very pleased that the company was doing well.

By the time I came to Caltech and realized the entrepreneurial landscape here, I think it had probably evolved to the point where faculty were very much encouraged to innovate and create, and Caltech is a partner in getting that technology out to people. I have to say very candidly I have a great relationship with Office of Tech Transfer. They funded our work on a number of occasions through their various funding programs, but they've really allowed me to decide what happens with that technology whether it's going to a company of mine or to an outside entity. I believe Tech Transfer's goal in operating this way is to hopefully help ensure the highest chance of success that a technology would actually help people someday. I've never had any issue in terms of what I wanted to do with that technology versus what Caltech wanted to do with that technology.

ZIERLER: Sarkis, what are some of the boundaries that you have to draw either financially, intellectually, even legally between your professor life and your startup life?

MAZMANIAN: This I think is true at any university. As we mentioned, at the end of the day, Caltech owns everything that we discover. If we were doing sponsored research with an outside company, let's say my company or any other company, what's really important is to make sure from Caltech's perspective that proper ownership to any invention is assigned. Meaning that if I innovated it it should belong to Caltech and not to the company that we were partnering with. That has changed over the years, David. I'd say because what has become harder to do here at Caltech is essentially to support R&D, research and development, for my own companies. I understand the premise that the motivation for that is that I have equity in a company. I stand to benefit from any financial rewards that that company may get from that technology, so there may be a temptation for me to make the data look better than it is so that it results in a financial outcome. Maybe that's a case by case basis, but I don't think faculty would ever think about anything for their own personal gains or certainly not to alter data in ways—maybe alter is not even the right word—but to frame data in ways that isn't reflective of what the data are actually telling us.

I think this is a concern for Caltech and I think it's a concern for the Office of the General Counsel. I know it's a concern for NIH, National Institutes of Health, which funds our work. Now you have this issue of Caltech ownership of the entity and their motivations, and I think the entities I work with are super collegial and they want to do the right things as well, even though they are financially motivated, and we are not. But also, the NIH which we're using taxpayer money to innovate. If we're using taxpayer money to innovate that shouldn't benefit a for-profit entity. I think that dynamic has caused Caltech—I can't speak about other institutions—to become very conservative in allowing me to collaborate with my own companies.

ZIERLER: A-ha.

MAZMANIAN: At this point we have almost, it's like the tail-end of a collaboration that's been going on for several years now. Likely the hurdles are just so high now to collaborate with my own companies that I just won't do that moving forward. I have to say I do think it's a shame. I understand the motivation as I said, but I do think it's a shame because we have not just a knowledge base, but a certain suite of critical technologies that you cannot replicate anywhere else. For the companies to really be able to maximize development of the technology, they should be able to leverage these unique capabilities that we have. Again, the world works a certain way and I guess I understand that.

ZIERLER: What about graduate students? How do they fit into the boundaries in terms of they're working here in a lab, but they're also looking for careers on the other side of their thesis?

MAZMANIAN: Again, there are firewalls between my students, all my trainees, and the company in any for-profit collaboration that we have and those are put in place through something called a research management plan that is required for any of these applications. But I do have students who are entrepreneurial themselves, and so I do want to expose them to industry so that that can be part of their training. This can happen through casual conversations and it does happen through casual conversations. The formal training with any industry partner—it doesn't in any way have to be an industry partner that I have equity in—happens when a student takes a formal leave of absence from Caltech and either interacts or works with that company. Many of my students have done this. There's a clear demarcation of when they stop innovating for Caltech and when they're innovating potentially for a company. My students have done this in the form of summer internships where they've really worked closely, in some cases they have actually been embedded in industry, they learned that landscape, that environment as well. In terms of my students or even any of my trainees learning from my interactions with industry whether it's my companies or others, that doesn't happen in real time.

ZIERLER: Sarkis, because so much of your research has medical applications, potential or real, is it a strategic blind spot that Caltech does not have a hospital or is that not really relevant for what you do?

MAZMANIAN: I used to think it's a blind spot, but now I think it's an advantage.

ZIERLER: [laugh] That's great.

MAZMANIAN: Because I'm becoming more mature, I guess, but Caltech knew this way before I started. Oddly enough I just came back two days ago from Washington University in St. Louis. It's a huge medical center that is home to some great basic science as well, but the environment is different, the culture is different. Everything is almost geared…the thinking is geared towards if you're not addressing human health then you're not doing worthwhile research. I think that's fundamentally flawed.

ZIERLER: Not to mention the fact that so much advancement comes from basic science for which we had no idea that it would have applications.

MAZMANIAN: That's right. All the stories are rich here at Caltech of people just following their intuition or their intellectual curiosities and not even knowing that there was translational potential to—

ZIERLER: Nobody talked about COVID when we discovered the mRNA in 1961. [laugh]

MAZMANIAN: Of course. These things evolve over time as well, but there's even more tangible examples of people again just doing fundamental research because they're curious. Then at some point either they or someone else realizes that there's an application for that. I think that there are only a handful of places like Caltech left in the world because there's just been so much of a push toward emphasizing medically-oriented research, and that comes from funding agencies, I think it comes from society as well. The approaches that we use, the value propositions that we have for our research as basic scientists, have evolved. The cost for that is that this blind, almost naïve exploration into the unknown which has yielded so many advances—not just for health, but just technology and innovation and quality of life—we're going to lose that. Sometimes you just have to go exploring without knowing why you're exploring, but you're just exploring. Caltech allows us that environment without this external pressure that you have to be thinking about an ultimate objective for your research. I think discovery for the sake of discovery is elegant and beautiful in addition to all the tangible outcomes. Again, as I get older and hopefully wiser I'm realizing just how valuable a place like Caltech is.

ZIERLER: I'm curious what your thoughts are on the perennial efforts to make Los Angeles a biotech hub—putting a hospital here, having a larger collaboration with UCLA and USC—if you support Caltech being more involved in these ways.

MAZMANIAN: I support it and I'm part of many of those activities as well. I think that's the way to do it as opposed to Caltech having a medical center, or medical school, or being affiliated closely with a medical center. I think at this point I would probably oppose that. Again, I can't believe I'm saying these words. Five years ago I would've never said that.

ZIERLER: [laugh]

MAZMANIAN: To do this through partnerships, to have conduits to interact with medical professionals, whether they're physicians, clinical trial specialists, people who have valuable human samples that we can access and do basic, basic research on—I think that's the way to do it because then you keep that mentality that we were just talking about. I don't mean to pick on Washington because I think that happens everywhere. You keep that mentality at an arm's length because then you can go back home and operate in a culture where discovery is revered. I think whether it's LA or Southern California or California or even not having these constraints being geographic, I think the access is valuable. But I think keeping Caltech insulated from those external pressures, that is something that again makes Caltech special.

ZIERLER: I asked about Caltech not having a hospital as a potential strategic blind spot. What about for you not going to medical school? Did you consider that at some point? Was human health always a motivating factor in your basic science?

MAZMANIAN: It came to me later. I became interested in basic research as an undergrad and then just followed a traditional academic path in basic research. Probably halfway through graduate school I realized that again, I can leverage these basic science findings to hopefully help people. It came to me later in my career. At that point, I wouldn't have turned back from a basic science path. Again, I'm glad that I did because I think it's a sign of the times because there are basic scientists and some people who just want to do fundamental discovery work and they never even want to think about application. I think that's beautiful, but again, I think you can do both. I think, ultimately, if we can help people why wouldn't we? At least that's my perspective and I think others who share that perspective or don't share that perspective, I respect them as well. For me maybe just another way to look at this, is at this point in my career I wouldn't be able to get as enthusiastic about projects that didn't have some translational potential down the road because again, I believe we can do both.

ZIERLER: What are some of the advances that happened more broadly in the field that allowed you to start thinking about human health not from the beginning, not at the early years of graduate training?

MAZMANIAN: Obviously advances in genomics, advances in technology, the microbiome space is a new field in and of itself. I think it was just more like an awakening for me. Just a realization that I didn't have to sell my soul if you will for gains to help people at the cost of doing fundamental discovery research. Again, I think it was more just a realization.

Evolution as Intellectual Framework

ZIERLER: Some overall questions about theoretical guidelines that you might rely on for your research. Evolution. Where is evolution as a framework for your work?

MAZMANIAN: We think about evolution, but it's hard to test evolutionary concepts especially when we're still trying to connect so many of the dots in what is a really new field. The microbiome, these bacteria that evolved inside of us, have become so specialized that they don't live anywhere else in ecology. They don't live in terrestrial or aquatic ecosystems. Many of them don't even live in other mammals. I think evolution has really formed this partnership between humans and microbes which are different than nonhuman primates and their microbes are different than other organisms down the evolutionary tree. It really speaks to just how closely intertwined we are with these organisms. What those forces were—again I can probably spend a lot of time thinking about that as well—are hard to define, but what I would say…and maybe this is going to be a small sliver of all the different evolutionary forces that go into building the microbiome.

Our work has actually shown that we're not hardwired to know the difference between friendly bacteria and harmful bacteria. Our immune system is geared to distinguish our cells from bacterial cells. Who's good and who's bad is hard for our immune system to know. Again, we don't have the receptors at the basic science level to know this is a bacteria that has potential infectious properties and this is a bacteria that isn't, so I'll welcome the hopefully friend organism and I'll reject the pathogenic organism. What we and subsequently others have shown is that the bacteria almost do the discrimination.

What bacteria that have evolved to live in us do, is they have changed those molecules which activate our immune system to become less prone to eliciting an immune response, but what's really interesting is that they've also developed new functions that suppress or down regulate our immune system. Not in a way that immunocompromises. In fact, quite the opposite; it actually bolsters our immune system. It prevents our immune system from attacking them. They can set up a harmonious partnership with us, a physical partnership with us. It's almost like a truce that they've forged over many millions of years. It makes sense that they're the ones who adapted, not us, because we have two copies of each gene so much harder to manipulate our genomes. When our genes mutate that leads to developmental issues or cancers, whereas bacteria, they're unicellular organisms who can take greater risk with mutations. They can just take chances and they can just rearrange their DNA and mutate their DNA and if it confers an advantage then those species persist and multiply. If it's deleterious then those organisms die. Again, they can take these evolutionary chances and I think they have.

What they've evolved are these pathways that essentially result in a "do not attack" signal sent to our immune system. You and I have a hundred trillion bacteria living in our colons, many of which are like snuggled right up against our immune cells. Our immune cells don't attack them. But you eat 10 cells of salmonella in contaminated chicken and your immune system fully wages war. The difference are the evolution of those bacteria. Where I think this has gone is that those initial encounters between us and microorganisms weren't always symbiotic or beneficial, but they learned to adapt and the organisms that live in us are no longer eliciting this immune response. I think the organisms that we consider pathogens—that are infectious agents in humans—are newly-evolved organisms. They just haven't learned to set up that truce. Maybe some will, some won't. The numbers bear this out. If we look at microbial strains within the human population, we associate with something on the order of 10,000 different microbial strains of which only 70 cause disease out of the 10,000. The vast majority of the organisms that we encounter are not infectious agents. Again, I think they've just learned over millions of years to no longer cause disease.

ZIERLER: Sarkis, when you're looking at a particular human health malady whether it's IBD or autism or Parkinson's, what are some of the through lines, the bigger takeaways, as they relate to age old debates about lifestyle choices, diseases of modernity, and simply the genetic hand that you've been dealt from your parents?

MAZMANIAN: I think they all play a role. Maybe I'll rephrase the question just to see if I understand it. What are the lifestyle or genetic factors that lead to microbial or microbial-induced maladies? Is that roughly how you're framing the question?

ZIERLER: Sure.

MAZMANIAN: The answer is in most cases we still don't know. We're still trying to understand the relationships in a way where we can predict is it some sort of predisposition to disease? That predisposition could be that I have a particular diet that fosters more inflammation in my body or more of an inflammatory microbiome and then that leads to some disease outcome. Or it could be that I live in an environment where there are particular toxins in the air or just elements in the food, in the environment that may dysregulate the system, our systems or theirs, that leads to disease. Certainly those are possible.

What I believe is this: I believe ultimately that there's already obviously some evidence for this, but I think ultimately the greatest driver is that within the human population there is a diversity of genetic predispositions that interact with the microbiome to impact disease. In other words, these are not disease genes, for example. This is just like normal polymorphisms, the changes that happen just through genetic drift that create the diversity of genomes throughout the human population. Some of us will have a genetic landscape that makes us more or less predisposed to the altered effects of a microbiome.

What I mean by that is, for example, I may have a gene or a mutation of polymorphism in a gene that in and of itself doesn't cause disease, but that makes me susceptible to some change in my microbiome and that change. It could be a new cell type or a change in the behavior of an existing cell type or a change in the location of an existing cell type. When that occurs maybe because of diet, maybe because I just encountered…I went to the wrong place at the wrong time—because I have this predisposed genetic landscape that that change then coupled with the genetic predisposition led to a manifestation that we would characterize as a disease. Again, if I have a headache that's not a disease, but if I have a brain tumor that's a disease. A lot of that is definitional.

Again, we've shown this in animal models and I think there's some correlative evidence in humans to suggest that that's the case. If this is true I think it's potentially leverage-able to help people because I don't think in most cases the microbiome or change in the microbiome directly cause disease. I think they're coupled with some other sensitivity or vulnerability, namely a genetic one. If that's the case that you indeed need a two-hit trigger for disease. It's very hard to change your genes. In 2022, it's very hard if you have a mutation that's associated with a disease—not causal, but associated with it—it's very hard through gene therapy to change that. But you sure as hell can change your microbiome. There's a number of ways you can do that, either non-invasively or subtlety through diet, even through like fecal transplants or probiotics or what have you. If you knew enough about the system where you removed the microbial trigger, the genetic trigger itself won't be enough to cause disease. I think this gives me a lot of hope that we may be able to address human diseases through this perspective.

ZIERLER: Your answer emphasizes how much we don't know, which is always the fact of the matter when you're working at the cutting edge. With that in mind and given the fact that as you well know that the microbiome, the gut health, it's in the zeitgeist right now. People are talking about it. There's a general awareness that this is a big deal. Where do you welcome that awareness and where is that problematic because of all of the wild claims out there about eat this—not that; do this—don't do that?

MAZMANIAN: There's a lot of hype. There's a lot of hype both in the public, meaning like the media and outside of scientific circles, but there is a lot of hype among scientists. I guess that happens. To some degree people want to promote their work. Maybe there's even…because that might lead to funding and what have you. I think we need to embrace our ignorance and realize that even though maybe intuitively one can see that there's huge potential in gut health of the microbiome—in the gut or other organ systems or other tissues as well—to helping people, I think again we don't know enough about many of those interactions, of many of those situations to be able to know how we can manipulate them or leverage them to help people.

Most of really the validated therapies or interventions are still very blunt. They're very crude. Antibiotics or wholesale microbiome changes. That just speaks to our ignorance. You and I each have somewhere between 200 and 500 different species of bacteria in our intestines. If I knew which bacteria was making what molecule that resulted in inflammation in my gut then I can target that, but we don't have that information. We're peeling the layers of the onion.

ZIERLER: That's a computational challenge, fundamentally?

MAZMANIAN: It's computational, it's biological, it's whether or not we're working with the right systems, it's ignorance. As I said I think there's a lot that goes into it. Again, from the company's standpoint just sequencing DNA and measuring all the molecules that you have doesn't tell you what those interactions look like. There's still a fundamental knowledge gap in terms of what are all the possible interactions, let alone what are the ones that are most relevant. There are other limitations as well. You and I are different. What may be causing disease in you would be different than what's causing disease in me. I don't always know where to look. What little nook and cranny in the intestine is that dysregulated interaction happening in that may be leading to a disease?

I think this just takes a lot of knowledge and research to start to understand what those interactions are. Oftentimes they're subtle. Not only are they personalized, but they're subtle. I think that this notion that you can just take a probiotic where there's a commercial too and it's going to solve everyone's problems…I think it's just naïve. Maybe it's just commercial; I'm not sure. And what their incentives are. But I don't think the systems work that way. I think they're very highly specialized, and until we have the knowledge to understand what those individual interactions are in any given person or place or time it's hard to think about what may work or what won't work.

These blunt interventions of, "I'm just going to remove all the bacteria." If the bacteria are triggering inflammation I'm just going to remove all the bacteria. I mean it works because you've removed all the bacteria, but what you've done is you've also removed all the beneficial outcomes of those organisms as well. If you just take broad spectrum antibiotics you might be removing that inciting agent, that organism that may be triggering inflammation, but you're also removing all the organisms that are producing cofactors and vitamins for you. You're also removing all of those organisms that are helping you digest your food, that are sending all these beneficial signals throughout your body that we believe exist, but we don't fully know.

It's not a long term solution. Or, if you just wholesale transplant microbiomes because there's something wrong with your microbiome, so I'm just going to replace it with another microbiome. Well, you're also replacing potentially beneficial interactions as well. Long term, we don't know if those fecal transplants as they're called have any side effects. The cases in which they've been used, specifically fecal transplants—some people call them FMTs, fecal microbiota transplants—are for very acute reactions, like some infectious gastroenteritis where one of the microbes in my gut is attacking me. I don't know which one it is, so I'm going to put out the fire by removing the microbes that I have and putting in all new microbes. That works in the short term, but does that long term cause other health issues, immunologic, metabolic, neurologic health issues? It's completely unknown. The short term gain of putting out the fire—in some cases you need to because some of these situations can be life or death and that I understand. But that short term gain may not balance out the long term harm that we might be doing. Again, it's still unknown because the first fecal transplants—at least clinical fecal transplants that were well documented—were only done in the past decade or so.

ZIERLER: Your answer hints at the underlying challenges of correlation versus causation.

MAZMANIAN: That's right.

ZIERLER: In the lab, intellectually, as a matter of intuition, how do you make those distinctions in coming to conclusions?

MAZMANIAN: I think that the causation argument is made oftentimes because more impactful or may get you more press or may get you more funding. I think those interactions between our microbiome and us are in most cases not causal. I think they go back to a lock and key relationship where there's something about us, our genetics or metabolism, that leads to whether or not that interaction is harmful or beneficial. I think outside of certain infections I personally don't believe that microbiome is causal to any human outcome. Let's say any disease outcome. For example, if it makes me more mentally clear that's one thing. Is it going to suppress a tumor? That's a whole different argument.

This is all theoretical, but let's say we can take the causality argument off the table. What you're left with are correlations or contributions, I would say. Correlations are things changing my microbiome, but those are changing as a reaction to something else. Maybe they're informative about what's changing them, so maybe my microbiome changes because my immune system is more activated. The microbiome is not causing my immune system to become more active or contributing to my immune system becoming more activated. It's just a read out or a consequence, a barometer of a more activated immune system. I think that's certainly possible. I think that happens a lot. There may be diagnostic potentials there, but I also do think that there are situations at least in animal models that we're quite confident about where we know the microbiome is actively contributing to outcomes, even disease outcomes. But again, not by itself. That animal, or I believe the human, has to have some other susceptibility or vulnerability where that alteration or that change in the microbiome is resulting in an outcome. I think that's sort of the spectrum of the potential interactions, I would say, of correlation, contribution, or actual causation.

ZIERLER: The answer also emphasizes whether it's diet or physically putting something in or taking something out of the body. What about other environmental changes that don't have that physicality to them such as stress or sleep? Have you looked at them and their potential impact on the microbiome?

MAZMANIAN: We are starting to look at stress; we have not looked at sleep. Others have been looking at stress as well and clearly again in animal systems and in some correlative studies in humans…clearly if you in animals induce stress you change their microbiome. In humans we don't do those intervention studies. We follow people…I guess in some cases we do, but they haven't been applied to the microbiome. Sometimes they'll mock up like an interview. [laugh] They'll put people in an arena or a situation where they think they're being interviewed, so I guess that inherently causes stress. What we do is we just generally track people over time. A lot of this is done in college students, so they'll look at college kids and sample their microbiomes when they're on spring break versus when they're taking their final exams. You can sort of model stressful situations. I think the data are pretty clear that the microbiome changes. You can correlate experiential factors with changes in the microbiome. What that means in terms of biology is a completely open question. Do those changes then manifest in some sort of feedback that have tangible effects on us? I've never seen a study that I would believe demonstrated that. It's certainly possible; I just don't have the data for it.

ZIERLER: If you could take me on a verbal tour of your lab—the specimens, the instrumentation, the tools—what's most important to you on a material level to make the discoveries you're working on?

MAZMANIAN: The mouse is the workhorse, for better or for worse. I mean the mouse is not representative of human disease and I think that's fine because again, we're exploring basic interactions between microbes and their hosts. There are plenty of microbes that live in mice that interact with their immune systems and other organ systems in their bodies, in their nervous system and metabolic system. I think if we had a core competency these days it's setting up animal models that tell us something about the microbiome's role in complex emotional behavior and neurodegeneration—the death of neurons in the brain. There are many other things about the microbiome that we cannot possibly study. In terms of the material…I guess I'll take a step back.

It's the students and the trainees. Ultimately, they're the ones who make magic, but I know you weren't asking about that. I think the other strength of our laboratory is the ability to measure behavior. How is an organism moving, making decisions, reacting to triggers or perceived stimuli and modifying its behaviors accordingly? I think we are pretty good at measuring the reaction of the mouse to some trigger whether it's microbial or otherwise. We also are pretty good at getting down to the cellular, molecular level…asking in those cases where we can flip a switch. An example being that we can change a microbiome in a number of different ways; that's a switch. Turn on the lightbulb, meaning, that we can change the behavior of that animal by flipping that switch. What is the circuit? What is the electric wire connecting the switch to the lightbulb? At the end of the day those are cells and molecules. We have a pretty good capability. Again, a lot of this is through collaboration to really get down to the cellular and molecular level and asking, "What cells in what particular brain regions are changing when I flip that switch?" What is that telling us about how those changes are resulting in a behavioral outcome?

The second part is hard because it's really hard to know how the behavior of a cell changes the behavior of an animal. I think the first half of the circuit is one that I think with some degree of confidence we can deconvolute in our experimental systems. This goes back to what a wonderful place Caltech is. We have one of the largest and I think one of the most capable gnotobiotic animal facilities in the country, maybe in the world in academia. A gnotobiotic facility is an animal facility where we can carefully control the microbes in the mice of the animals that are in our animal colonies.

ZIERLER: Through what mechanisms?

MAZMANIAN: We keep mice in sterile bubbles. They are essentially these plastic chambers where all the air is HEPA filtered in and out. All the food and bedding and water is sterilized, all the instruments are sterilized. We can exquisitely control what microbes these animals are exposed to. We can put in hundreds of microbes, we can put in one microbe, we can have mice that are probably germ free.

ZIERLER: It's a perfectly controlled space, you're saying?

MAZMANIAN: Perfectly controlled space. It allows us to do experiments where we can get down to single organisms and single pathways and really take the system apart. If you look at the basic problem as we have, mice have many hundreds of bacterial types in their intestines. How do you ever know which organism or which group of organisms matter in any context? Or maybe it's like all organisms because it's general microbial stimuli. You've got to be able to add and subtract bacteria. To do that effectively and rigorously you have to have a very controlled system. Many years ago Caltech very generously built us this incredible gnotobiotic facility. I think it's really supercharged our work. In many ways they built a facility bigger than what I ever asked of Caltech. [laugh] Again, that speaks to how unique Caltech is in really enabling laboratories to reach for the stars and to do experiments that are very challenging to do in many other environments. That's been super enabling for our work as well. Otherwise I think the instrumentation in our laboratory is pretty basic. I wouldn't say we have any one instrument that is unique or rare that we leverage to add to our research. Ultimately, it's cell and molecular biology using the unique capabilities that we can generate with animals.

ZIERLER: You mention that it's all about the students and the trainees. That's where the magic happens. That's perfect because for the last part of our talk today I'd like to focus on some general questions about your work as a mentor to students both at the undergraduate, graduate, and postdoc level. For undergraduates, their sensibilities, obviously their skills with computers, what are some of the things that they're interested in now that might be modern, that might be new and innovative relative to you when you were an undergraduate or even the beginning of your career as a professor?

MAZMANIAN: It's coding. That's the first thing that comes to mind. The ability to write particular scripts or programs that allow students to analyze datasets of interest in their projects. This is something that I certainly wasn't exposed to and it's still magic to me. The students, especially at Caltech, they all are very computationally savvy. I think that's the perspective they bring and the graduate students to be frank as well. I think that's a perspective they bring that I think is more a sign of the times in training these days as opposed to even 10 or 15 years ago.

ZIERLER: The fact that computer science is overwhelmingly the most popular major among undergraduates, is that an asset to you? Do you get the follow on benefits of that for students who are in computer science, but they have biological interests?

MAZMANIAN: I think the undergrads—if we're speaking only about undergrads—who gravitate to our work are the ones who are interested in biology and in particular, are interested in human health. For all the reasons that we've talked about of what makes Caltech special is there isn't this huge focus on human health across many of the laboratories. I think that makes certain laboratories more attractive to those individuals who may be considering careers in medicine or just have life experiences that make them want to study particular diseases or just have their own intrinsic interests in health-related problems. I think that's what attracts the overwhelming number of undergrads who come to our lab because they're interested in some aspect of human health. There's too many other avenues at Caltech to at least directly explore. As we talked about there are many indirect ways of addressing human health, but we do it relatively speaking more directly.

ZIERLER: In your interaction with undergraduates, when you have the opportunity to give some advice about next steps when you see their skillsets when do you encourage graduate school? When do you encourage medicine? When do you encourage industry?

MAZMANIAN: I wouldn't say I ever encourage career paths. I try and be responsive to where the students want to go. I'll do two things. I'll answer their questions from whatever perspective that I can give them. Second, once they have a focus for their careers, I can help equip them with the tools, essentially the knowledge that they need to then execute. I'm fairly careful—I think it comes naturally to not impose my thoughts on any of my trainees. I'm here to help them achieve their career goals. I think just in these conversations I naturally almost watch myself sometimes. If anything I've prevented myself from saying things to students because I think that it may overly influence their decision making.

ZIERLER: You said earlier for your graduate students—almost jokingly—they're teaching you as much as you're teaching them.

MAZMANIAN: I'm not joking. [laugh] It's absolutely true.

ZIERLER: What are the areas where graduate students are coming in with a stronger knowledge base than what you might have?

MAZMANIAN: It's computer science obviously, as we spoke, but then also neuroscience. All the problems that we're addressing in our laboratory broadly speaking could fall in—and maybe that's why we're in the building as we talked about—broadly can fall into the large umbrella of neuroscience. I'm completely ignorant about neuroscience. I've read one textbook my entire life, a number of original research articles, but have no formal training in neuroscience. Really a lot of my understanding of neuroscience comes through interactions with students who both have formal training in neuroscience or are learning neuroscience as well, but they're just much more invested and they have the bandwidth to do that. Those interactions, those conversations teach me. I guess I have institutional knowledge in immunology and certainly in microbiology as well. I'm really proud of the fact that the lab is interdisciplinary. What I mean by that is not that immunologists come in and they contribute immunology to their project or train others or help others with their immunology programs. They come in with the formal training in immunology and they learn neuroscience as well. They learn metabolism as well. They learn computer science as well. I think that when they leave they're just much more eclectic in their skillset in a way that I think I rationalize as being better able to understand all the interactions that are happening in a complex organism like a mouse or a human.

ZIERLER: To the extent that the thesis research of your graduate students now really represents the vanguard, the frontier of where the field is headed, what are some of the big question marks? Where are some of the places they're going that might suggest where the discipline is headed in the next five or ten years?

MAZMANIAN: There's a number of ways to answer your question, but I keep coming back to the human health perspective. I hope the more immediate gains will be in the space of neurodegeneration and not in terms of behavior. I say that because behavior is just so complex. The fields of neurobiology and neuroscience are still trying to understand the cellular and molecular interactions that lead to changes in behavior. We can certainly look at whether a cell is dead or not. I think, based on the fact that with a great deal of confidence understanding the endpoint—is a cell is alive or is it dead—gives us the ability to then go back and say, "What is the process? What were those interactions that led to that particular outcome?"

In a mouse, David, I'm convinced that the microbiome modulates both the death of neurons as well as the replication, neurogenesis, the birth of new neurons. I think the evidence is crystal clear, to be as honest with you that you can—using these very blunt approaches—that I can remove the microbiome of a mouse and see changes in neurodegeneration or neurogenesis. Where we are now is: what are those specific interactions? What are those specific pathways that lead from the gut to the brain that regulate the death or rebirth of neurons? I think that is a problem because I think fundamentally it's a chemistry problem. These molecular signals are sent from microbes to our own cells. Maybe not directly to the neuron, maybe to some other cells in our body. It could be an epithelial cell in the gut, it could be an immune cell, it could be a neuron in our gut that then leads to signals in the brain. That is a problem that we should be able to start unpacking and start resolving. I'm hopeful that the short term gains—short term in biology is five to ten years—would come in the space of neurodegeneration.

ZIERLER: Finally, last question for today just as a capsule, a snapshot in time. What are you currently working on?

MAZMANIAN: We are working on a number of different model systems. Roughly speaking, we work on anxiety behaviors in mice, we work on autism like behaviors in mice, we work on Parkinson's disease. A new project that's just getting off the ground, we work on Alzheimer's disease. We don't develop new mouse models. That's sort of a principle of our lab because we want to use, we want to make rapid gains, so we use mouse models that have been validated, well established by the field. What we bring is this new perspective that events in the brain may be controlled by events in the gut. Our technology and our knowledge base allows us to ask those questions and begin to answer them.

ZIERLER: Sarkis, this has been a great conversation. Thank you so much for spending the time. Next time we'll go back all the way to the beginning. We'll learn about your childhood and your educational trajectory.

MAZMANIAN: Thanks, David.

[End of Recording]

ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Tuesday, June 28th, 2022. I'm delighted to be back with Professor Sarkis Mazmanian. Sarkis, it's great to be with you once again. Thanks for joining me.

MAZMANIAN: Thank you for having me, David.

Armenian Origins and Cultural Commitments

ZIERLER: Sarkis, in our first conversation we did a great overview, a tour of your approach to the research, the big questions in the field. Today I want to start by developing your personal narrative and family background. Let's start first with your name. Both your first name and your surname—what is the national origin and what is the meaning of either your first or last name?

MAZMANIAN: The origin is Armenian. According to 23andMe I'm 99.3% Armenian as defined by the region where my grandparents and great grandparents came from. I'm happy to talk about the geography there as well. The name…I don't know what any symbolism or meaning to my last name is, but Sarkis is an Armenian saint. There's a church in Pasadena called Saint Sarkis Church. It's actually serendipitously where I got married. It wasn't because it was the saint that I was named after, but that's the origin.

ZIERLER: Tell me about the geography, where your family comes from going back a few generations.

MAZMANIAN: I traced my grandparents birthplace back on my father's side to a city named Akn, but pronounced AGEN in Turkey. On my mother's side—I shouldn't even say city; it's sort of like villages—to a village named Antyeb. I know this because my parents have told me where my grandparents have come from. There's a history that goes back to the Armenian genocide that started in earnest in 1915 and continued for many years afterwards where that area, which is part of historic Armenia and people had lived on that land, essentially the ancestors of our Armenian people were forced to leave their city, their homeland for a variety of reasons. Mostly because they were being persecuted for their religious and ethnic beliefs. Not to get too political or historical about this, but that's where my family to the best of my knowledge, where my grandparents were born and where my ancestors had lived for thousands of years prior to that. My grandparents moved at a very early age to Syria where my parents were born. Then they later moved to Lebanon and I was born in Lebanon and then moved to the United States when I was very young.

ZIERLER: Was the move to the Middle East, to Syria and then Lebanon, was that originally prompted by the genocide in Turkey?

MAZMANIAN: It was. This is just history where Armenian civilizations were taken into a desert, essentially marched through the desert. The name of the desert in northern Syria, southern Turkey is called Der Zor. We have poems and songs about that desert and what Armenians did along the way. Many of them perished along the way and some historians have concluded that was the intent of marching people through the desert. Along the way, Armenians formed villages and ultimately cities as well and then settled in that country, in Syria, settled in that space. None of this was by choice, maybe just to put it directly.

ZIERLER: This exodus, did this take place on both sides of your family?

MAZMANIAN: It did. All four of my grandparents were survivors of the Armenian genocide.

ZIERLER: Where did your parents meet?

MAZMANIAN: My parents met in Beirut. Many Armenians who ultimately wound up settling in Syria started moving to Lebanon in the mid-50s because the climate in Syria had turned a little bit against Christian Armenians and Lebanon is just a more tolerant country. It always has been one of the most tolerant countries in the Middle East. It's very welcoming and accepting, so they moved to Lebanon, obviously independently of each other. They met there in 1969 and then got married in 1970. Most Armenians who lived in Syria had moved to Lebanon by that time.

ZIERLER: Was there a large expat Armenian community in Beirut?

MAZMANIAN: Very large right up to the point where over the years I've just met so many Lebanese people who have told me that they just grew up with Armenians. There are Armenians in the neighborhood and pools and they always talk about the other friends that they knew from childhood who were Armenian and talk about food and the Armenian culture they were exposed to. In many ways, I don't know the exact statistics, but Armenians were probably one of the largest ethnic populations at least in Beirut, in the capital of Lebanon, to the point where most Lebanese people grew up with neighbors or knowing and having friends who were Armenian.

ZIERLER: How did Armenians in Lebanon deal with larger geopolitical issues? For example the wars with Israel?

MAZMANIAN: I don't know. It's not something I've explored. My father would have known. He passed away a year ago. His whole life he was very much into following politics. He never was a politician himself. It wasn't his nature to get involved, but it's something he was quite interested in. This is something that he probably would've had some insight into, but again not discussions I would've had with him. I just haven't read enough of the history to know where Armenians sort of stood on the broader issues in the Middle East with Lebanon and Israel and other areas.

ZIERLER: For your parents and the larger Armenian community, how well assimilated were they into Lebanese society?

MAZMANIAN: I would say very well assimilated to the point where we exchanged language. Part of the Arabic language is now part of the Armenian language and I suspect it went in the other direction to some degree, at least in that particular region itself. The food, the culture, the music. This was the way my parents had described their lifestyles in Lebanon, as if they were living seamlessly among the native Lebanese people. Very much to this day actually. Very much a productive part of society and an accepted part of society. Very well integrated into Lebanese culture. The Armenian population has dwindled over the past several decades in Lebanon, but I still have a first cousin who lives in Lebanon. I have very few other relatives and they're more distant than my cousin. The way he tells me is there's largely no distinction between Lebanese and Armenian people. Again, at least in the large city.

ZIERLER: What level of education did your parents achieve and what were their professions in Lebanon?

MAZMANIAN: My parents, neither one of them have any education. What I mean by that is they never went even to elementary school. Growing up in Syria there were Armenian schools, but those were for people who had means, and also people who didn't have to work themselves. My parents both effectively had to work from a young age. My father's first job began when he was seven years old. He was sweeping hair in a barbershop. He worked his whole life after that. Never went to school, though I think if he had gone to school, I think he'd had been quite successful especially in an academic environment. Whether or not that would translate into some profession I'm not sure. He was an avid reader and maybe one of those metrics is that by the time he passed away he was fluently speaking, reading, and writing six languages.

ZIERLER: Wow.

MAZMANIAN: Again, never had a formal education. The way he would phrase it to me was that while other kids were kicking a ball he was reading a book.

ZIERLER: A-ha.

MAZMANIAN: That's the sort of person he was. My mother stayed home and helped with the house. She had two brothers who were from a young age working to support the family as well. They just never had the opportunity to go to school. One of the things that my father mentioned to me, many years ago now, was the first time he saw the inside of a classroom was my older sister's parent-teacher night when she was in first grade. So, no education at all.

ZIERLER: Sarkis, what year were you born?

MAZMANIAN: 1972.

ZIERLER: Where was it? What was the neighborhood?

MAZMANIAN: I was born in Beirut. The name of the neighborhood was called BourjHammoud. I can't tell you much more than the name of the neighborhood. I don't remember. I was a year and a half when we came to the U.S. and I've never been back. I don't know much about the neighborhood itself.

ZIERLER: Obviously you wouldn't have known this at the time, but what languages were spoken in your family's house circa 1972?

MAZMANIAN: Armenian, Arabic, Turkish were the languages that were spoken. My father also spoke French at the time. My mother didn't speak French. Unless he wanted her to not understand something when he was talking to somebody else. As you may know, Lebanon was a French colony, so there were a lot of French speaking people in Lebanon. My mother wasn't able to speak French. Those three languages, most prominently Armenian and less common Turkish and Arabic.

ZIERLER: What were the circumstances of your family moving to the United States?

MAZMANIAN: This actually partly relates to what I mentioned that my father followed politics. In 1975 there was a civil war in Beirut which essentially changed the course of history for that city. It's never recovered from that war. We moved to the U.S. in May of 1974. My father was so convinced that there was going to be a war approximately a year before it started that he left his business and two homes and essentially took my mom and the three of us—I have two sisters—and moved us to a foreign country where he knew one person—which was my uncle on my mom's side actually—and left everything behind. From what he tells me his family and friends all thought he was crazy for doing what he was doing, but he was right in the end. He wanted to get us out before it was too late.

The circumstances were that he…I think there's one event, but he's never…I want to be careful because I don't want to put words in his mouth, but he's never linked these two events. There was one thing that must have affected his thinking. He was shot eight times several months after I was born. He was just in the wrong place at the wrong time. It was politically motivated. One group was after another. He was in a coffee shop or a tavern when he shouldn't have been and was caught in the crossfire. He lived most of his life with five bullets in him. They removed three. Again, I suspect that had a lot to do with his desire to leave Lebanon. [laugh] Again, enough of a clue that that world was imminent.

We were fortunate enough to have the financial resources to hire a lawyer and get the right documentation and come. He's told me that the last hurdle was the people he had to bribe at the airport even though we had all the paperwork. We needed to leave the country. Ultimately the business he had, the homes he had, everyone lost everything. He would've lost everything anyway. For many years afterwards he worked to get many of my other relatives, both on my mom's side and my dad's side, here as well. He was able to successfully help them through either by sponsoring them to come to the U.S. or just helping them navigate the legal aspects of this. He came and built a life here again under what I think is a pretty unusual set of decisions which I'm not sure I'm bold enough to have made. [laugh]

ZIERLER: Where was your mother's brother living at that point?

MAZMANIAN: In Los Angeles. That's why we moved to Los Angeles, particularly in the Mid-Wilshire area.

ZIERLER: Was there already a large Armenian community in L.A. at that point?

MAZMANIAN: There were certainly Armenians here. I won't be able to quote the actual number of people or percent of the population, but I'm confident there were at least a couple of Armenian schools and churches here already. There must've been a population of several thousand at that point. Most of the Armenians have come to Los Angeles—there's about half a million Armenians in Southern California now, most of which came in two waves. One was the wave that we were the harbingers of and that's from Lebanon in the ‘70s, and then another wave in the early ‘90s that came mostly from Armenia. That's because of the fall of the Soviet Union. Armenia was one of the 15 for Soviet republics and so when the Soviet Union collapsed there was a lot of turmoil and many Armenians came at that time as well. Yes, so this is all the last two or three generations.

ZIERLER: Sarkis, what was your earliest memory in Los Angeles and perhaps a recognition that your family were not native to the United States?

MAZMANIAN: My earliest memory… I can place a memory to a time when I must have been probably three years old. Whether or not there are other memories that I've had over the years I'm not sure. It's not going to make me look good. I'm going to keep this short. My mother was learning how to drive. In fact my uncle was teaching her how to drive and I was so upset that she left that I kicked a hole in the door after she left me. Again, I remember the house we were in and there was a hole in the door when we left. It was a hollow front door which, now in retrospect… I'm not sure was the right structural design. [laugh] That was my earliest memory. In terms of us being not native I'm not sure I can recall either an event or a feeling or thought related to that to be honest with you. We went to Armenian school so maybe that colored my view of where I fit into society. I can't recall anything at least off the top of my head in terms of our ethnic role or differences in being in Los Angeles.

ZIERLER: Where in L.A. did your family first land? What neighborhood was it?

MAZMANIAN: It was Mid City, so we lived on Olympic near La Brea for a couple of years and then moved to Hollywood when I was probably around four years old. Then grew up in Hollywood until I was 13. Then we moved to the San Fernando Valley after that.

ZIERLER: Did your father's brother-in-law also have a job waiting for him or how did he start making money when he got here?

MAZMANIAN: I doubt he had a job waiting for him, but he was a car mechanic. That's something he had learned in Beirut. When we arrived he was working at a mechanic shop. Later he had his own auto-body mechanic shop. For most of his life he just operated a gas station here in Pasadena actually. Interestingly enough the same uncle passed away three weeks ago. His funeral was last week.

ZIERLER: Oh, wow. I'm sorry.

MAZMANIAN: I think we were scheduled…were we scheduled to speak on the 15th?

ZIERLER: That must have been the reschedule.

MAZMANIAN: That was the day of his funeral.

ZIERLER: Oh, wow. Wow. Sarkis, what was your first language?

MAZMANIAN: Armenian. We spoke Armenian at home. Before I went to school I was speaking Armenian. From what I understand from what my mom told me that I learned English watching TV and then obviously continued to. I went to a private Armenian school most of my life and I guess learned a little bit of Spanish along the way, but really I'm mostly conversing in Armenian and English to this day.

ZIERLER: To the extent that scientific abilities have a genetic or at least a familial origin is there anybody in your family that you can trace that had a special ability in math or science?

MAZMANIAN: Nothing formal. I have a sister who was born a year older than me and she was the first in our family to go to college. By profession or trade no one who would be in STEM that we define today, but certain qualities…I think about my mother more so than my father. She was much more analytical in her thinking and very good. She handled all the finances in our house. Just knowing her math came quite easily to her. One example would be…one of the things that was really remarkable about her is that we'd go to the grocery store and she'd keep a running tab of how much each item cost that she would put into the cart. This became almost like a game between me and my sisters of how close could we predict the answer because she got to the register and she'd tell us what the cost of all the items were and she'd be quite close. We were just always trying to compete with her as well.

ZIERLER: [laugh] When the family moved to Hollywood was that part of a larger upwardly mobile Armenian move once they got settled and spent a few years in L.A.?

MAZMANIAN: I think so. A lot of this is in retrospect for me because I wasn't aware of what was happening. The story is not too dissimilar from many other Armenians who had come at the time and I'm sure many other people as well that obviously I'm just not familiar with. Whether it's our immediate family, those who I know most about, and look at their trajectories over those years and of course, all the stories I've heard. Again, this is likely a very biased statement, but I think Armenians are just hard working people and understand certain aspects of business. It's one of the things that I think my father learned in Beirut and probably was just part of the cultural environment that he was in. I think that led to the ability to go from, in my father's case, being an employee at a small business when he first started to then a few years later buying that business from the owner who he worked for. Then expanding that business to the point where he was able to buy a house, an apartment building, as an investment in Hollywood. We lived in one of the units of the apartment building. Then he was able to buy a few years later a single family home and I guess live the American dream. Really just based on the fact that he just came and did something related to his previous profession, but was able to just learn new aspects of a business to the point where again he was able to purchase that business.

ZIERLER: Sarkis, I wonder if you have a sense if the two Armenian communities in Los Angeles—the one coming from the former Soviet Union and the one coming from Lebanon—if they reconnected somewhat seamlessly or there were divisions or distinctions among the communities.

MAZMANIAN: There are distinctions to this day. Similar to many other ethnic groups there's differences in some of the foods and the traditions, even the dialects that we speak. There is some level of labeling going on of where people are from and so we have particular names for whether or not you're from Beirut or from Armenia or from Iran. There's a large population of Iranian Armenians or Turkish Armenians as well, so there's names. The history of this, David, goes back to really traditional Armenia, back 2,000-3,000 years or more even where we have written language where historic Armenia was called the Armenian Highlands as well, were composed of something on the order of 13 different cities. It's actually very similar to Greek history where there were essentially city-states and each of these city-states operated by themselves even though there was sort of an ethnic glue that held them all together. There would be wars and disputes between each of these cities or regions, but when there was an external invader they would all come together and hopefully fight off—in some cases lose to that external invader. This identity issue is something that is historic in Armenia, but also the fact that we can identify, that people are from a different group, doesn't mean that we don't still feel close. I think that most Armenians I suspect would share this view that regardless of where in Armenia they come from, and we can maybe poke fun at it and bring it up from time to time. They behave this way because they're from a different region than I am, but at the end of the day we still feel close to that person even though they're a little bit different from us. Again, I suspect a lot of that goes back to ancient history.

ZIERLER: Sarkis, in your family what were some of the big traditions or foods or holidays, cultural markers that stick out in your memory?

MAZMANIAN: They were really events that brought the family together. To be perfectly frank obviously we'd celebrate birthdays and holidays, but it wasn't so much the tradition or the religion or the event. It was really these were opportunities for us to come together as a family. I can say this because a vast majority of the time that we'd spend let's say at Christmas or Easter were not ceremonial or in any way sort of honoring a tradition. We were just catching up and spending time with each other. My parents were religious, but I wouldn't say overly religious. We'd go to church for important events or important holidays.

ZIERLER: This was an Eastern Orthodox church?

MAZMANIAN: It is. To be specific it's Armenian Apostolic, but it's essentially Armenian Orthodox. Very similar to Russian and Greek Orthodox. They were really opportunities just to come together as a family. Food was very traditional Armenian food. In fact, I didn't have a hamburger until I was 18 years old.

ZIERLER: [laugh]

MAZMANIAN: I wouldn't order pizza until I was in college just because my mom would just cook and we'd eat Armenian food at school.

ZIERLER: Which you didn't call Armenian food. You just called it food. [laugh]

MAZMANIAN: Right. I'm joking because obviously they'd serve hot dogs and hamburgers at school as well, but it was a little different. In no way did my parents try to instill a sense of nationalism or culture in us. It was just the way their daily lives were and the decisions that they made were always going back to their traditions. It wasn't that they weren't open to incorporating other traditions or cultures or ways of life. It's that they really weren't exposed to it. These were my younger ages so it's not as if they wouldn't have made Chinese food; they just never saw or came across Chinese food themselves.

ZIERLER: What do you think your parents' motivations were in sending you to Armenian school and not public school?

MAZMANIAN: Very much I think this is cultural as well. Just being in a different country is to make sure that we're maintaining the culture, the language, the traditions. It was something that was very important to them. It's still important to me now. I have two young kids, five and seven, and they go to Armenian school for the same reasons. It's just to make sure they're being exposed to being Armenian. What's different with my kids is they're also exposed to many different cultures in ways that I was not when I was younger. My parents' motivation really was essentially just to keep those traditions, the language alive being in a different country.

ZIERLER: Not that you would've been able to compare it to public school, but was science and math strong in Armenian schools when you were growing up?

MAZMANIAN: Not particularly to the best of my knowledge. In fact, part of this was that it wasn't emphasized in the school. Part of this was I didn't have much of an interest in science growing up actually all the way through high school. If there was an emphasis I probably wasn't paying too much attention to that as well. I would say in general that there wasn't. Really the emphasis was more on history, on language—both Armenian and English, and social studies. Less so on the sciences. I don't know why that was. I think it goes beyond my perception. If that was just how the school was organized or that they just couldn't recruit science and math teachers. I just don't know.

ZIERLER: Growing up how insular was your world? Did you hang out, did you play with non-Armenian kids?

MAZMANIAN: We did not. In addition to our school which was very homogenous the neighborhood that we lived in…there were many Armenians. Not exclusively, obviously. I think that my parents probably felt most comfortable for us to play or visit other families that they knew. The world was small. I think both the time plus also the ways they sort of engaged and connected with society…it's not like they were part of organizations and groups that were more community based and less ethnically based. I think the fact that all of their circle was Armenians we were only exposed to Armenians. Whatever interactions I had with non-Armenian people at a young age were very casual and not deep friendships.

ZIERLER: Did you go to Armenian school through high school until you graduated?

MAZMANIAN: I went to Armenian school until ninth grade. Then I actually went to a public school after that. This was something my sister conjured up at some point was that she convinced me to essentially lobby our parents to put us into a public school. I have to say just to be honest it was a great decision, but I didn't understand the wisdom of that decision at the time. What I told my parents with the coaching of my sister was that we wanted to be exposed to different cultures. We wanted to see what the real world was like and being in this Armenian bubble was something we wanted to break out of. The lever that my sister pulled to make this happen was that we had moved from Hollywood to the San Fernando Valley to Van Nuys. It was quite a long commute because the Armenian school was in Hollywood. It was quite a long commute for my dad. My sister said, "Why don't we just walk to the public school two blocks away and you won't have to drive us across town and have to pay for that education?" It wasn't a difficult decision for my parents. I think to their credit they always let us make important decisions and let us take the lead on certain things. I think once they saw that my older sister and I wanted to essentially go to school in a very diverse environment they were happy to have us make that decision.

ZIERLER: What high school did you go to?

MAZMANIAN: Ulysses S. Grant High School.

ZIERLER: How much of a culture shock was that for you?

MAZMANIAN: You know, David, it actually wasn't. It's surprising that it wasn't because I went from a small parochial school to one of the largest schools in the Los Angeles Unified School District. I think it was 3,500 students at that time from everywhere in the world. I don't remember it being a difficult transition to be honest with you. I wasn't a particularly social kid either, so it is in retrospect surprising that it wasn't more of a difficult transition. I remember making friends pretty quickly. I remember feeling comfortable pretty quickly. In fact, I don't remember a period where I regretted transitioning to a public school. I'm struggling to find the factors of why that acclimation was seamless, but it was.

ZIERLER: Sarkis, in public school did this wake up latent interests in math and science that you might not have been able to pursue in parochial school?

MAZMANIAN: It did not. In fact I took one science class in high school. It was chemistry. I took it in the summer just to get it over with quickly. If anything I had an aversion to science.

ZIERLER: Wow! [laugh] So, you were not a science whiz kid in high school?
MAZMANIAN: No. I was not. In fact I spent a lot of time writing in high school. If I was ever passionate about anything in high school it was really about writing, mostly poems and short stories. I'd always do well in class and submit my poems and short stories to various different contests both in school and outside of school. I remember that being what I was mostly interested in. I started off college as an English major. Science wasn't even on my radar. [laugh]

ZIERLER: That's amazing. How big a purview did you have in terms of places to apply? Would your parents accept it if you went to school outside of Los Angeles?

MAZMANIAN: They would've. My parents were pretty progressive at the time compared to I think other Armenian or maybe just other immigrants. As I described they lived in a bubble. They hopefully felt that was safe for them and they probably did, but to venture out of that was probably if they thought about it—I'm sure they did—was unsettling or maybe even terrifying. But they let us make our own decisions. My parents have always been…if we went wanted to do something they would always tell us their opinion, they would always tell us their thoughts, but they would always let us make our own decisions. I remember from even the youngest ages that I can recall where I had to make a decision and maybe they didn't agree with it, they never stopped us from doing anything we really wanted to do. So, I applied pretty broadly to college. I had no desire to go far anyway. I applied to several of the University of California schools. The only school I applied to outside of California was Harvard. I didn't get into Harvard, but I got into a few UC schools.

ZIERLER: Obviously Caltech would not have been on your radar if you were interested in pursuing English as an undergraduate.

MAZMANIAN: I probably shouldn't say this, but in high school I didn't know what Caltech was. [laugh]

ZIERLER: [laugh]

MAZMANIAN: I was only about 15 miles from here. A hardcore science and technology school was not somewhere where I would have applied regardless. I just wasn't looking for that as an environment for me for training. In fact, to be honest to myself I'm not sure I knew what I wanted to do. I wasn't sure I wanted to even go to college for a while. Not that I had other career plans or some other trajectory, I didn't know how to make those particular decisions. They weren't as important to me as they were for other kids. I remember being in high school and explicitly being in conversations where everyone was talking about college and I'm wondering why is everyone talking about college? It just wasn't something that was important to me. Maybe it had to do…I don't know if it's just the conversation now. Maybe I had to deal with the fact that my parents never went to school and so it wasn't something that was ever prioritized to us. My parents never asked me even once if I was going to college or where I was going to go to college. It was both a combination of it's something that they were unfamiliar with and they just trusted us to make those decisions. To what degree the decision of college was something that other kids were focused on because of their parents or because that was just what they were seeing other people talk about or if it was something they were really thinking carefully about in terms of their next steps, whatever those influences were I don't think I had those same influences even from my peers. As my peers were talking about college it was not a topic that I really wanted to engage in. I applied to college because it was the thing to do. It wasn't my goal, but the fact is, David, I had no goals. [laugh] It's not as if it was at the expense of something else.

ZIERLER: Did you know enough, Sarkis, at least that you did not want to pursue a working class life like your father?

MAZMANIAN: I think that I probably would've been happy at the time pursuing a working class life. Not that I had a goal, but I was working…I'm kind of bringing this full circle. I was working in a body shop in high school. We had something called the work study program that allowed me to get out after two-thirds of the day and spend the last part of the day in a work environment. I was working in a body shop for a car dealership in Van Nuys. This was my last year of high school and I thought to myself I guess I can do this for a living. I hadn't decided to do it for a living. I just was sort of figuring things out, but I had no aversion to that. Not seriously, but even whenever it entered my mind I'd think about maybe a career as a car mechanic or repairing cars. I had no issues with that whatsoever.

ZIERLER: Here must be the big narrative turning point. You come to UCLA intending to study English. What turned you on to science or who turned you on to science?

MAZMANIAN: I guess it was a series of events. Part of the reason why I chose English was at the advice of my teachers in high school. I realized that just because I guess I had a talent for it didn't mean it was something that I wanted to do. It was really in my first year of college where I started thinking about my future. I was fortunate enough to have good enough grades to be able to get into college, so I was an undergrad at UCLA. Once I started the classes I just realized that I'm not sure I wanted a career in areas that an English degree would allow. I started thinking more along the lines of…or just being more introspective in asking what are the things that I want to do with the rest of my life, not what are the disciplines that I want to study. It really was kind of this journey where I started almost searching within my first and transitioning into my second year of college. I started searching for something that really resonated with me, something that really made me excited. I actually switched to being an economics major for a couple of terms because I thought maybe that would be a segue to business, but didn't really like economics for a number of reasons. In particular was that I felt like the vast majority of economists were practicing theory and not developing new theory. It's like 1% of economists would develop new theory and those were the ones who win the Nobel Prizes. It wasn't creative enough for me, whereas English was quite creative. I guess I was looking for something that allowed me to have that originality and that creativity.

In my second year I took my first undergrad general elective science class. I guess this is reflective on that I didn't like science since I pushed this off to my second year, but maybe I matured or maybe I was just in the right frame of mind. It was an undergraduate intro biology class where I guess for the first time the way I would describe it is that for the first time I wanted to see how the story ended. I was reading about the phylogeny of plants or something. I found myself wanting to turn the page.

ZIERLER: Like it's a story essentially?

MAZMANIAN: Yes. I wanted to learn. I was probably 18 or 19 years old and for the first time I was fascinated by something I was reading in a book. It just snowballed from there, David. I hadn't fully decided that I wanted a career in science, but I just started taking more and more science classes. By the end of my second year I had made the decision that I would switch to a science based department or major. I hadn't taken any microbiology classes at the time, but I started talking to school counselors and to other students. I had sort of put the pieces together that if creativity is driving me then the way that I would be able to satisfy that urge to do things creatively was to study a discipline that was very research oriented. At least at UCLA it's the way it was taught. You're obviously creating when you're doing research and I just learned that microbiology was one of the most research oriented undergraduate majors within the life sciences at UCLA. In fact, I was a microbiology molecular genetics major. I switched at the end of my second year and then continued and then got my degree in microbiology and molecular genetics.

Most of my classes were research paper based and much less memorizing particular discoveries. More along the lines of analyzing the experiments and the thought process that went into those papers and those discoveries. Interestingly I was clearly floating around for many years of my life, but once I found microbiology I've actually never looked back and never done anything else. I do other things with our research, but my formal training was always in microbiology and obviously a huge aspect of our research is microbiology now. I've never looked back and thought I could have or should have done something different.

ZIERLER: Sarkis, a broader political and sociological question. When you were at UCLA as an undergraduate…the beating of Rodney King and the L.A. riots in 1992—was that a big deal in your life? Was that something that you talked about with your family as they tried to understand what was happening?

MAZMANIAN: To be honest, David, I don't think I had the context to fully grasp what was happening with racial issues, policing issues. Obviously I was aware of what was happening; I can tell you where I was. When the L.A. riots happened—I believe it was in '93 or '94. I could be getting that wrong. It was during that timeframe. I think that's right. I was both based on my upbringing, but also the fact that I was just so focused on my work that I was probably…a lot was happening in the world around me that I was just simply not grasping. I remember the details, but I know that I wasn't immersed in that cause and really understanding what the roots of those issues were. It's almost embarrassing now to think back and look at that, but I do believe that that is accurate.

Jumping Into Science at UCLA

ZIERLER: Sure. Sarkis, as an undergraduate what were the most important laboratory classes for your education?

MAZMANIAN: I loved organic chemistry. [laugh] I'm not sure I remember most of it, but I really enjoyed both the quantitative aspect and essentially the way it's just like a big puzzle. The way organic chemistry works, at least my naïve understanding is that you have all these different functional groups and all these different chemical structures and you think about interesting ways of putting them together and developing new molecules. I thought that was just absolutely fascinating at the time. I remember some of the other lab classes, especially those that were more biologically oriented where we were doing simple dissections. I liked them, but they certainly didn't excite me as much as thinking about chemical reactions and then going into the laboratory and actually doing those reactions and seeing if I got the product that I was hoping to generate. It was probably my first and last exposure to organic chemistry, but I remember really being fond of it.

ZIERLER: Did you work in science in the summers? Did you stay on campus to work in the labs?

MAZMANIAN: I did both throughout the academic year and during the summers. Once I switched over to a microbiology major—I guess it was the summer between my second and third year. I was in college for five years because all of this moving around meant that I had to take more classes to satisfy the degree requirements. I was essentially in my second year starting my first year from a unit standpoint. Once I decided to become a microbiology major I applied for a work position in a laboratory of someone who was affiliated with the microbiology department. He's still at UCLA; his name is Jake Lusis. He's a colleague of mine now. I applied for a job in his lab, and he gave the job to somebody else, but introduced me to a colleague of his who was in the Department of Pathology and Laboratory Medicine who was looking for help. He said that she had an opening. Her name was Judith Berliner and she offered me a job.

I went and the job was essentially to wash dishes in the lab. I said to her, "I'm happy to wash dishes, but can I do some research as well?" Very quickly I was doing research working under the postdoc essentially doing atherosclerosis research. It was very cell culture based, so it was pretty amenable for an undergrad. That's what I did for the next three years was work in Judy's lab trying to understand how specific immune cells called macrophages enter into aortic tissue and become these own cells which are part of the development of atherosclerotic plaque. It was cell culture based research. I didn't discover much; I remember a lot of failed experiments. [laugh] I just remember being fascinated by research. I knew I wanted to do research for the rest of my life after that. There was this set of serendipitous events that got me into Judy's lab that was really formative and very helpful even though I didn't accomplish much as an undergraduate researcher.

ZIERLER: Sarkis, in Judy's lab was there any component dedicated to translational research or this was all fundamental science?

MAZMANIAN: It was fundamental science, but everything was in the context of translational research because the singular focus of the laboratory was atherosclerosis and heart disease, essentially. There was definitely a translation component to it, but like many things now the actual approach to understanding how one can improve human health is to really reduce it down to very simple and basic components which are far away from even the human being, let alone the disease, and just really manipulating or exploring hypotheses in reductionist systems. In her case it was endothelial cell lines and macrophage cell lines.

ZIERLER: What were some of the considerations in terms of staying at UCLA for graduate school? Generally the mode is you go elsewhere for graduate school, broaden your wings a bit.

MAZMANIAN: I knew I wanted to take a year off while I was applying to graduate school. After I graduated I applied for a research technician position to pay the bills. It happened to be, I suspect I searched for positions in microbiology because I was most familiar with that, and I was able to get a research technician job in a microbiology lab at UCLA. That's why I continued. Then I was applying to graduate school at the time. Again applied to most UCs and UCLA as well and then was accepted to UCLA. I'd grown interested in the research I was doing as a technician that a large part of my decision to remain at UCLA for graduate school was because I was pretty intent on joining that same laboratory for my graduate studies. I ultimately did. I went through the whole process. I actually left the lab for six months, did rotations in other laboratories, but then came back to that same lab where I started as a research technician and that's where I got my Ph.D.

ZIERLER: In graduate school what did you see as the skills that you were developing that would be most relevant and useful as you were thinking about a specialty and topics to pursue for thesis research?

MAZMANIAN: I have to give credit to my mentor. His name was Olaf Schneewind. He passed away a few years ago. He instilled this…almost a compassion to be able to design an experiment that very rigorously told you an answer to a very specific question. I think this is in many ways a lost art these days of just the way we do science. We cover so much space now with these approaches that allow us to generate huge amounts of data. At the time at least the focus and the training that I got was the opposite end of the spectrum. How do you reduce a problem down to a very, very, very simple question that's a component of the larger problem and then design an experiment that is unequivocally going to answer that question and not some other related or unrelated question? I just remember sitting there for hours and just thinking about the design of an experiment. I guess I figured I can execute the experiment, but how do you design an experiment controlling for all of those variables where I'm going to get the answer—positive or negative, whatever it may be—but I'm going to answer this question with this experiment? I was very focused on building that skillset of just being an experimentalist. To this day I think that that's still…again we don't do science generally that way anymore because we just have tools and approaches that allow us to cover so much more space, but I think it's such an important skill.

If you're designing experiments where you can have multiple interpretations or they can give you results that are not fully able to answer your question—maybe answer a related question or even lead you into directions that are unanticipated—then that wasn't the best experiment. That wasn't the most optimal experiment because it didn't move your project forward in the most efficient way. I think this is something I try and teach my students. I'm sure I don't do as effective a job that Olaf did with me, but that was really my focus where I was able to answer questions. Later on after three or four years I started thinking more broadly about what were the important questions. Perhaps by then I had more confidence so I could answer them if I could identify what those important questions were. I started thinking and reading more broadly as I expanded my horizons and started thinking about what are the areas of science that excite me and interest me?

The Vanguard of Bacterial Research in Graduate School

ZIERLER: As you articulated Olaf's special skill, what were some of those specific questions that he postulated for which he developed these very specific experiments? What were the big questions that he was after?

MAZMANIAN: Technically…what was the research focus? Olaf was an assistant professor when I started. I was his third graduate student. When I joined as a technician there was only one graduate student at a time in the lab. I was very junior which to me was really beneficial because every day was a huge learning opportunity because he was always in the lab and always ready to tell me what he thought and share his opinions. What Olaf had discovered a few years prior to me joining the laboratory when he was a postdoc or a fellow at Rockefeller University was that there are a specific subset of proteins which are found on the surface of gram-positive bacteria.

Just really briefly in terms of microbiology there are two general class of bacteria. This is really an unfair simplification, but this is how it's actually taught. There are gram-negative and gram-positive bacteria. Gram-positive bacteria, they have a particular structure on their surface called peptidoglycan—otherwise known as the cell wall—which essentially gives bacteria its rigidity and its structural integrity. It's almost like a shell around the bacteria that protect everything that's on the inside—the DNA, the proteins, the membranes, what have you. If this shell is what is exposed to the surface and the environment of that bacterium how does bacterium interact with its environment? How does it communicate with its environment? How does it manipulate its environment? Part of the way it did so were by covalently anchoring proteins to the peptidoglycan, to this cell wall.

What Olaf had discovered was that this process happens. He had shown that there were bacterial proteins covalently anchored to the cell wall which is a major discovery which he made in '93. I think that's when that paper was published. When I joined the lab he was on the hunt for the enzyme that performed this reaction. He hypothesized that indeed if these bacterial proteins were being anchored to the cell wall in this particular way which requires a chemical linkage, an enzyme must be performing that reaction…was the most obvious explanation for it. What I did when I joined the laboratory was to set up a genetic screen to identify the gene that encoded for the enzyme which performed this reaction.

The importance, the relevance of this question was that many of the bacteria that he had shown that have these proteins on the cell wall were pathogens, were bacteria that caused infectious disease. For example, the bacteria that cause strep throat or group B strep that you may know of, staph aureus, staph epidermidis, even some forms of microbacterium which are related to gram-positive bacteria is all human pathogens that have these proteins on their cell wall. The assumption at the time which one of my later projects proved at least in mice was that these proteins, part of what they're doing was to promote the infectious process. If we can discover the enzyme that links these proteins which help the bacteria become more infectious then we may be able to drug that enzyme and make the bacteria less infectious if that makes sense. That was the context and motivation for the research program.

What I did in staph, aureus which causes staph infections which can affect many, many different parts of the body was all in vitro—all on microbiological plates—was to set up a genetic screen where I one by one mutated the genome or the DNA of staph aureus and then asked, "Can I find particular strains or clones of bacteria that no longer perform this reaction?" I was trying to essentially mutate or reduce the activity of this enzyme by altering the DNA that encoded for that enzyme. In bacteria you can do these large screens because you can make many, many mutants of bacteria. You can make a thousand mutants of bacteria and if you have a very quick assay to know have you disturbed a particular biological function you can say this clone of bacteria is missing a function that this other one has, the parent organism or the wildtype organism. If it's missing this reaction and it's missing it because I mutated its DNA then where I mutated the DNA must be a region of importance that is linked to this activity I'm interested in. Maybe it's a convoluted way of saying that, but hopefully that makes sense. That's what I did.

I spent a year setting up a genetic screen to find staph aureus mutants that no longer anchored proteins to the cell wall and I was able to do that. Then I spent another year—and this was before the age where you can just sequence the DNA of bacteria which takes an afternoon now. It took me a year just to sequence one little region of that bacterial genome, but I did all manually. Not using robots, but sitting with radioactive isotopes and labelling the bacteria and using very basic molecular biology techniques. I found the region that was mutated; I found that that region was within a gene and encoded for what we now know was that enzyme that anchored these virulence factors to the gram-positive cell wall. That was my first major discovery in science.

ZIERLER: Sarkis, of course as a graduate student you're hyper focused in the laboratory. To what extent were you aware of all of the developments in the field more generally? I'm thinking about the Human Genome Project, the biotech revolution, the understanding that DNA sequencing was going to lead to new and amazing things. Were you following these trend lines at all as a graduate student?

MAZMANIAN: I was more so, I would say, in the subsequent years as opposed to in graduate school, where again my horizons really started to expand, but I was very much aware of it. Thank you for this discussion, David, because you're forcing me to go back in time in ways that I have really not done in the particular perspective that I hadn't appreciated yet. What I realized is that the first few years of graduate school I was very, very focused on not just microbiology but the specific field or subspecialty of microbiology that involved gram-positive bacteria, and these biosynthetic pathways to anchor proteins to the cell wall. Over time I suspect I exhausted that literature because I was reading quite avidly and started layering on new aspects of microbiology. These were the days, David, where—I was describing this to a graduate student last week and they thought it was quite fascinating. Actually, I should ask them what they really thought. These were not the days where you can just go online and pull down a manuscript of interest. These were days where we didn't have computers.

ZIERLER: Right. Go to the library.

MAZMANIAN: You had to go to the library and search through manual catalogs of tables of reference, a table of contents. Then when you wanted a journal you had to go in the book stacks and find the volume that had all the different issues of that particular journal that you're interested in and find the page and photocopy it and bring it back to the lab and read. It wasn't as easy to do your literature work; it took a lot of time. I really very much enjoyed reading quite broadly towards the later years, three, four, five of graduate school where I started venturing into other areas of microbiology. I guess towards the end of graduate school outside of microbiology. It was really during my postdoctoral years where I started thinking about genomics and health and immunology and biochemistry and a little bit of neuroscience at the time. I can't say that I was exposed to much of what was going on in industry or biotech. That came later for me.

At the time I thought to myself, this was the culture of academia. There was really a huge barrier, an almost firewall between academia and industry. There were very few private collaborations at the time. Science is very different now than it was back then. Maybe it was just the circles I was in, but I think it was pretty pervasive. Academics stayed in their own circles and industry stayed in theirs and there was very little overlap. I know that there are some reasons for this. Just very briefly, industry had large R&D operations where they would make their internal discoveries largely within the company. That whole model has changed in the last 20 years where there's a lot more collaboration with academia and academia is essentially doing a lot of the R&D for industry.

With that said I wasn't exposed to much of industry, but I would say…this was the year 2001-2002 where I started reading about—at the time you couldn't avoid reading about the Human Genome Project because it was so important. But I started thinking more broadly about eukaryotic systems, about animals and animal models, about to whatever degree I was exposed to about human health and all the different organ systems and cell types and interactions that were at play in any complex organism. I guess that aspect of my training is embedded in almost everything we do now. I try and think about the complexity of biological systems and all the different biological systems themselves, but all the different disciplines that we have which focus on those biological systems and how they all need to be married and really brought together. Otherwise you're just looking at a biological problem in a vacuum. For example, if immunologists are really only trained to think about immunology they may be able to solve problems in immunology, but they're not going to understand how the organism works until you understand metabolism and neuroscience and endocrinology and all of the other what we view as disciplines that go on to explain these biologies in a complex organism. A lot of that started happening for me in my postdoctoral years.

ZIERLER: Sarkis, what were the most important instruments or microscopy for your thesis research?

MAZMANIAN: It was very much molecular biology driven. The instruments were very rudimentary compared to what we use now, but a lot of different types of gels to essentially resolve proteins or DNA from each other. Obviously this was really when molecular biology was gaining its footing and becoming so pervasive in many aspects of science where we can again manipulate DNA in ways that we couldn't even five or ten years prior to me starting graduate school. The tools were again cutting and pasting DNA both as a research tool, but also to understand biological systems. Then for graduate school coming back to having manipulated the DNA in bacteria, asking what did those changes mean to the biology of the organism or the specific process within the organism that I was interested in. A lot of the tools or the readouts that we used at the time were to see have we interrupted or augmented or changed a biological process within a cell by maneuvering its DNA around?

Some of the other tools that I also used was to then take those organisms that I made that either had a different context in terms of anchoring proteins to the cell wall, maybe do it less sufficiently, in some cases do it more efficiently and then ask what does that mean for the fitness of the bacteria in its ability to cause infection? We'd take some of those organisms and put them into mouse models and then ask did the bacteria cause different levels of the disease or different types of disease based on how I've manipulated this one biological process within the bacteria that gives you insight into what that process means to the microorganism itself? Basically many of the tools that molecular biologists were using at the time were in my toolkit.

ZIERLER: Sarkis, to go back to the admiring way you talked about the precision of Olaf's questions and the way that he created experiments to pursue answers to those questions. For your own thesis research how did you know you had enough to defend? What was that threshold for you?

MAZMANIAN: A lot of this came down…one of the ways to phrase this is the most important currency in biology is publications, I would say. As some metric, in some cases imprecise, of quantifying your level of training. I had a number of…you asked about my perception. Let me tell you my thesis committee's perception. I had published a number of papers and there's actually a plot twist a little bit. I had published a number of papers halfway through my fourth year of graduate school, so three and a half years into graduate school. The plot twist was that Olaf was recruited to The University of Chicago to be chair of microbiology. He was an assistant professor going to be a chair at The University of Chicago.

ZIERLER: He's a rising star at this point.

MAZMANIAN: That's right. He asked me to join him and move to Chicago. I presented this to my thesis committee and after three and a half years the feedback from my thesis committee was why are going to move to Chicago? Just graduate and go on with your life. You have enough here to graduate and defend your thesis. I'd made a number of discoveries that I think were pretty important and moved the field along. Publication is a way to quantify it, but that's really the essence of what they're looking for. It was a combination of the fact that I was very loyal to Olaf and he had asked me to set up the lab because he was taking on this administrative role and he asked me to essentially be the person on the ground setting up a new laboratory, but also I had made a new discovery at the time that I felt that I could really unpack in a year or a year and a half. I was really excited about this new discovery and I just didn't want to walk away at that moment because again the lab was going to move and leave. So, I moved to Chicago for a year.

To answer your question I'm not sure…I think…actually I know my thesis committee was ready to graduate me and for my thesis. I wasn't ready and so I actually stayed on another year and a half. I have to say this was quite a learning lesson in terms of how do you build a career? I learned a lot in that next year both in terms of the technical aspects which I think are secondary to how do you ask and answer research questions? This relates to what we were talking about a few minutes ago. I felt like I was really operating at quite a high, mature level as a scientist because I was able to put those pieces together. I put theoretical and more conceptual pieces and the actual practical pieces to help solve those problems. I was very efficient and I think it was almost a validation, almost a training ground to see can I put the conceptual and the practical aspects together and I was able to. I published a couple of nice papers in that last year at graduate school. I also distinctly remember that there was a period towards the end where I felt like the learning stopped. I felt like I wasn't learning anything more.

ZIERLER: Because you were at the vanguard of the field—is that what it was about?

MAZMANIAN: I felt like I could just do this over and over with new…within my little sphere in microbiology I could always find interesting questions and figure out how to answer them. I became quite confidant in that. That to me was when I felt like the learning stopped. It wasn't because I couldn't just dissect and explore a new biological system within bacteria. I felt like I needed to do something entirely different than what I was doing because I could just replicate my graduate career over and over and over. I felt I could do that, but that doesn't mean that I was learning anything new. I might learn something new about that particular question, but I wasn't learning anything new about how do you approach science and what this whole scientific world was outside of microbiology. If that makes sense. I just remember being at a point in graduate school where I felt like I just needed to move on. This is a little over a year after my thesis committee felt like I was ready to move on. I wasn't ready when they were ready, but I did become ready sometime afterwards.

ZIERLER: Sarkis, last question for today. What were the principle conclusions of your thesis research or what did you feel like you had contributed to the field more generally up to that point?

MAZMANIAN: The answer is already embedded in some of the things I've said to you. It's that there is a biological process, a particular dedicated process in gram-positive bacteria that is mediated by a single enzyme, quite a remarkable little enzyme that displays these virulence factors, these proteins on the surface of gram-positive bacteria. What we learned was that this process is quite universal to gram-positive bacteria meaning that there are dozens—and this has been shown in the years later—and dozens of other gram-positive bacteria that use the same pathway to essentially elaborate these proteins on their cell surface and really decorate what they look like to the outside world. Meaning the immune system, for example, or the sewer that they were in and they use these proteins to essentially perform their biological activity.

In the case of staph aureus what we learned was that this process was required for the billion bacteria to cause infections in animal models and ultimately a couple of the last pieces of discoveries were that we were able to drug this enzyme with a small molecule that inhibited the ability of this bacterial enzyme to anchor on the cell wall and when we gave that inhibitor to mice that were infected with staph aureus we were able to reduce the virulence of the bacteria. Essentially it was a novel anti-infective. Not antibiotic because it didn't kill the bacteria; the bacteria lived just fine. They just weren't able to cause disease in mice. In fact the drug company licensed the technology and they were going to develop and they ultimately didn't and that played out for another decade after graduate school. It looks like it was a viable target for a new class of molecules. It's not currently being developed now.

The other thing I appreciated—and this was one of my last discoveries that I made in Chicago—was that the same system had been coopted by the bacteria to essentially bring nutrients into the cell. Maybe I was very myopic early on thinking that a lot of these proteins were causing infection, but part of the infection process that I realized was that the bacteria had to be metabolically active, had to have the nutrients they needed in environments that oftentimes can be nutrient limiting. If the bacteria had evolved a process to essentially manipulate events at its bacterial surface, part of that was attaching to tissue, part of that was evading the immune system. That's the part that I was initially focused on, but my last two papers were that this same system also helped bacteria import nutrients into the cell so it could grow. It was a pretty versatile system and my graduate career took a trajectory where the research question guided me to thinking about nutritional importance during infection.

ZIERLER: Sarkis, on that note for next time we'll pick up on these new directions after graduate school, during your years in Chicago, and ultimately what got you to Caltech.

[End of Recording]

ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Thursday, July 14th, 2022. I'm delighted to be back once again with Professor Sarkis Mazmanian. Sarkis, once again it's great to be with you. Thank you for joining me.

MAZMANIAN: Thank you for having me, David.

New Directions at Harvard Medical School

ZIERLER: Sarkis, we're going to pick up right where we left off from our last conversation—the new discoveries and directions that you saw your thesis research taking you in. To set the stage for your next chapter did you see certain opportunities in the translational realm even that early that might have influenced the kinds of postdocs or faculty positions you would have been considering after defending the thesis?

MAZMANIAN: Yes. It actually wasn't the translational component that drew me to the particular topic area in the laboratory where I continued my career, but more this…I wouldn't say it's serendipitous, but it was likely just a set of thoughts that led me to understand that the microbiome was there. Everyone knew about it, but no one was studying this. There were really at the time—this was 2001—only a handful of laboratories around the world that were truly studying the microbiome. Even then, the way they were approaching the research was looking at one organism at a time, so a model organism or one microbe that was prominent in the human microbiome and studying the effects of that bug in isolation, not the entire complex community. When I defended my thesis, or what I was getting close to defending…I loved microbiology as we discussed and wanted to stay close to microbes, but also wanted to do something different. At the time most microbiology was focused on either infectious disease or metabolism. Obviously, there was some environmental microbiology as well, which has also exploded in ways that are really, really fun to watch. But really nothing else. I found that I predicted that the human microbiome was a new frontier, and I couldn't imagine how fast it was going to grow, but it did in the couple of decades afterwards and I just wanted to do something different.

I actually remember a specific moment where I realized that this was going to be an intriguing area of research. I forget what parts of this we had discussed, but I remember telling you that I moved to Chicago to finish up my thesis. I would take the train from Hyde Park where the University of Chicago was, up to downtown to The Loop to spend Sundays there and just get out of the little bubble of Hyde Park. I remember sitting on the steps of the Hancock Tower and just reading papers and I came across a one page editorial from a professor named Jeffrey Gordon, who is clearly one of the pioneers in modern microbiome research. Essentially the take-home message was what I just mentioned: that we know these microbes exist; we've known for hundreds of years these microbes exist in our gut, in our oral cavity, in our skin. But they weren't well catalogued and we knew very, very little about what they were doing. That alone was enough to intrigue me to jump in, into the unknown.

I formed a few hypotheses—I can get into them if they're of interest to you—of particular types of interactions or relationships between microbes and their hosts that intrigued me. What I did to identify an environment where I would be able to study this was I went through the literature and I asked, "What are the laboratories that have the systems or the model organisms or the perspectives that would allow me to go and to begin a new line of research." I wound up just applying to a handful of laboratories. My first interview was with Dennis Kasper at Harvard Medical School. Dennis is essentially an infectious disease doctor. He was chief of residents of infectious disease at the Brigham & Women's Hospital for many years. His specialty was in microbial carbohydrates. He'd been studying these sugar complexes on the surface of gut bacteria in the context of infection because we do live in a precarious relationship with our microbiomes. Certain organisms do cause disease. Most of them I believe are benign or beneficial. He had that concept of just thinking about the gut from an orthogonal perspective which as I mentioned is known as post-traditional from an infectious disease standpoint. Long story short—I can get into even more details; I think there's maybe even an interesting story there about Dennis embedded into my transition to his lab. I interviewed and after spending half a day with him I decided that's where I was going to do my postdoc. Actually it was a dinner plus a half a day. Then I cancelled all my subsequent interviews and did my postdoc there.

ZIERLER: Clearly the personalities as well as the science clicked.

MAZMANIAN: That's exactly right. I'll make this brief. Again, it tells you a little bit about Dennis, who is a wonderful man. The first was the dinner. I was telling him about my thoughts about the microbiome and specifically a project I had in mind which I'd never wound up studying until I started my own laboratory. The concept that raises the question, "Why do we immunologically tolerate so many foreign organisms in our gut?" When we are colonized with salmonella or E. coli, our immune system knows to become activated and to try and clear or control that organism, but the rest of our lives we're living with hundreds and hundreds of bacteria that we're not reacting to. We're not mounting an immune response. Why is that? Not that we have a clear answer even today, but at the time there was almost no evidence or no data to suggest one or another mechanism. That was a general concept that I was interested in. I remember clearly—and we've joked about this since then—his response was, "Sarkis, I have no idea what you're talking about, but you can come here and do whatever you want."

ZIERLER: [laugh] That's great.

MAZMANIAN: That was music to my ears. That was the scientific part. The personality part—that's why I said a half a day—the next day was my interview. I had meetings in the morning and then we went to lunch. On the way back from lunch we went up the elevator and then we walked out of the elevator and right in front of him was his receptionist to the department that he was running. As he was walking by he looked at her and something caught his eye and he was pretty disturbed by what he saw. He immediately asked her to walk in…there was an accountant's office right next door. He asked the accountant to walk out. He took the receptionist into the office and he started examining her because he's a medical doctor. He completely forgot about me. He completely forgot that I was standing there and he was trying to recruit me the whole time. He poked his head out and said, "Sarkis, can you give me a few more minutes?" The fact that he was so concerned that he saw something in her that made him forget about me and focus on her just told me all I needed to know.

ZIERLER: That's great.

MAZMANIAN: It turned out she had a brain tumor. She was fine. When I went into my postdoc she was the receptionist the whole time and she was great. We became really good friends. It was that connection that made me realize that he was who I wanted to work with. Again, it was a no brainer to not interview with anyone else.

ZIERLER: Sarkis, I want to get to those hypotheses that you were thinking about around those times. As a way of providing some context I'm not sure if you've ever looked into the history of the field, but the fact that these microbes were known for hundreds of years but they were not well catalogued…what's the bigger takeaway there? What was the story that previous generations of scientists had missed?

MAZMANIAN: I don't know if they had missed anything, but it's sort of like someone or some group of people have to be intrepid enough to start digging in a particular area and then martial the interest and the resources, the tools, the approaches for others to continue and for fields to grow. It was a convergence of technology. Prior to this time it was very, very hard to sequence DNA. You know all about this. We're talking about 20-plus years ago. That's actually the time when the Human Genome Project was being completed. What the Human Genome Project did was advance sequencing technology so that they could sequence the human genome. There was a lot of technology development in the decade prior to that. What people had done in terms of the microbiome, the very little that had been done, was taking single organisms cultured out of the gut and then studying them in a laboratory or in mice. But very hard to get a glimpse, a lens into what was happening in the intestines themselves. It was really DNA sequencing technology that broke the field open. I think in many ways it was likely there was just not enough interest. People just thought they're digesting certain carbohydrates, making a few vitamins and cofactors—made in the microbiome, but maybe there wasn't more beyond that. Again, you don't know until you study it.

I think it was really application of sequencing technology that got more and more people interested in the microbiome. That just happened to happen—it was not by any design—to coincide with when I was starting my postdoc. I just happened to be in the right field at the right time and then by the time I was done in 2006, there were a number of papers coming out on sequencing the microbiome and that's really how the field took off. The first set of really impactful papers were two human stool donors' microbiomes were sequenced, which made it to the cover of the New York Times. Now we do thousands in an afternoon.

ZIERLER: [laugh] Now let's go to these hypotheses that you were developing. What were some of the initial interests where you saw the research going from that point?

MAZMANIAN: The main one was what I just described. Why do we tolerate these organisms? Why don't we react against them? The other was how do we assemble and build these communities of organisms because we are all born microbiologically sterile? In the womb we have no microbes on any of our externally exposed surfaces, but within three to four years of life we know that we're teeming with hundreds of bacteria in our gut and our skin and oral cavity, vaginal cavity, and in some cases the lung as well. My question was, how does this even happen? What is the succession of events? What are the microbe interactions that leads to both shaping the gut environment to be more hospitable to colonization, and how do particular microbes welcome or repel other organisms?

You arrive at a configuration of many hundreds of species of bacteria, many trillions of cells that form some sort of stable and resilient community. That was really interesting to me. And one of the other things I was thinking about was on a very reductionist level – taking mice that had no bacteria in their intestines and then adding organisms one by one and looking at microbe interactions and how those interactions shaped the environment, which in my particular case, was the mouse, and looking specifically at the immune system. Again, I had this hypothesis about the immunologic tolerance against microbes. It was both on the microbiome side and how that assembles, but also how that impacts the immune system.

ZIERLER: I want to square the circle between going back to our original point in today's conversation about not being so mindful of translational applications at that point and yet being attracted to work as a postdoc with a medical doctor. Were you looking out in the far term where that was where this research was ultimately headed?

MAZMANIAN: I feel like I always had my eye on translation, but the individual experiments, hypotheses, and projects weren't set up initially to lead to breakthroughs that may impact human health. Again, it was very much basic science questions, fundamental research in these two areas of community building and its effect on the immune system. When you're talking about the immune system there's a natural progression or a natural translation to health. If I may then take this conversation and maybe fast forward and describe what happened in my postdoc. What I did was I set up a very old technology called gnotobiology, which is the ability to manipulate the microbes in an organism. The organism of choice for me was the mouse. For the first time in 40 years prior to that they had a facility at Harvard that would house sterile mice, mice that had no microbes. The facility had been dormant since then, and I was reactivating it. You'd put these mice in these flexible film isolators, where the air is HEPA filtered in and out, all the food in the environment is completely sterilized, and you can keep them and raise them without any microbes. You have to supplement their food with certain vitamins and nutrients, but the mice will stay alive.

I set up this facility, which took me the first year of my postdoc just to do that. I wasn't collecting any data. The idea was that now I can, with a lot of precision, control what microbes were in the animals so that I could do my experiments. Again, adding the succession of different microbes and then looking at the effects on the immune system. What happened was, as I was doing these experiments and putting various different microbes into mice in different orders and different sets of complexities, was that some of the first microbes I put in I just started profiling the immune system of those animals. A lot of this was because I was just trying to learn immunology because I had no training in immunology as a graduate student. I was just trying to learn the basic approaches, the basic techniques to profiling the immune system of the mouse. I made a pretty interesting discovery.

When I compared mice that you would just buy from the animal vendors that had dozens of bacteria—these are the mice that we use in the laboratory all the time—to mice that I had now raised myself that had no bacteria, I found that in the spleen—so outside the gut—in the circulation of those animals there was a huge difference in the proportion of T cells. I thought to myself, Why would an animal that has no microbes, the genetics are identical between…and the food…everything is identical between the mice that have microbes and mice that don't—why would there be a difference in their T cell compartment in the spleen? Even in the gut, it would be fascinating, even outside the body or outside the gut. I became really fascinated with this question of, Do microbes educate or prime the immune system in ways that make the immune system function normally? It's almost like turning our view of microbes 180 degrees because I think you and I and most people were raised thinking that we have an immune system to fight off microbes, that they become activated to control microbes. That's true for infectious diseases, but for the organisms that live inside of us it turns out that there are particular organisms which allow our immune systems to function in ways that we consider normal.

Again, everyone who had been studying these immune profiles for many decades just assumed that that was just because of the genetics, the DNA of the animal. Ultimately, the culmination of my postdoc was that we discovered that not only are microbes required for our immune systems to look and function normally, but I identified a particular microbe and a particular molecule from the microbe that was required for augmenting the proportion of T cells that were lost in the germ-free mouse. This turned out to be a major finding because again, both in terms of the concept that our microbiomes educate our immune systems and increase immunologic function, but also that there are specific bacteria with specific molecules that are required for this beneficial outcome. These were the major results of my postdoctoral work.

Now to tie to the translational part, I asked myself, Because I had discovered this effect particularly on T cells I started thinking are there T cell mediated diseases which then would be affected by the microbiome? In other words, if this increase in the proportion of T cells somehow modulated disease, can I set up experimental systems whereby adding these particular bacteria that affect the T cell repertoire can I either worsen or mitigate the outcome of an immune and specifically a T cell mediated disease? I did two things. One of the first things I did when I started at Caltech was both to understand what type of T cell was being affected by that molecule because all T cells are not the same. It turns out we have T cells that attack. The ones that we think about, but we also have T cells that are like the brakes on the immune system. If you think about the immune system, it's like a loaded gun ready to fire when we're infected, but in the absence of an infection we want to restrain or control the immune system. We have other T cells that do that. They're called regulatory T cells. These are suppressive T cells. It turned out that the type of T cell that was being affected most by this microbe were these regulatory T cells. I said, if this bacteria is inducing an anti-inflammatory response then I would predict that the mice that didn't have the bacteria would result in inflammation and that this organism would be able to suppress that inflammation if that makes sense.

I stayed close to the gut this time. I chose a model of inflammatory bowel disease, a mouse model of Crohn's disease. Crohn's disease is very similar—it's not an autoimmune disease—but it's very similar to autoimmune diseases in the fact that if you go back to this loaded gun analogy, that if the immune system is not restrained, is not kept in check, then the overall activation of the immune system leads to essentially the immune system attacking us. The immune system may think that there's a microbe or an infection when there isn't and that prolonged inflammation leads to tissue damage. In Crohn's disease that tissue damage is in the gut. Again, because of activated T cells which were not being suppressed. In diseases like multiple sclerosis that's in the brain, and diseases like rheumatoid arthritis that's in the joint – it's all the same concept that is this lack of regulatory T cells. You remove the regulatory T cells that pro-inflammatory T cells attack. They're attacking us and not the bacteria. That's really the basis of autoimmunity and allergic reactions as well, this whole concept that the immune system has to be kept in balance; it has to be kept in check.

I set up these models of Crohn's disease where I now added back this bacteria and showed that I was able to induce the anti-inflammatory T cells and that led to an amelioration of disease. The mice that had this particular bacteria did much better than the mice that didn't have the bacteria. We showed that the bacteria required this particular molecule for protection from disease and over the many years, the first five to seven years of my lab here at Caltech, we dissected this pathway, identified receptors, immunologic responses, how those T cells were protecting against intestinal inflammation. Just to maybe cap the translational part of this, then I started a company around this technology to develop this molecule as a novel biologic anti-inflammatory for both Crohn's disease, and then later we showed it actually works in mouse models of multiple sclerosis as well. There's a drug company now developing this technology, so maybe someday all this basic research that I just described to you may lead to a pill with either bacteria or a molecule from bacteria being the drug for Crohn's disease, multiple sclerosis, and maybe potentially other autoimmune diseases as well.

ZIERLER: Sarkis, to give some context to the research community around this time during your postdoc. While you're in discovery mode, what aspects of your research are you doing purely on your own and where is there sort of a multiple independent discovery going on more generally in the field where you're collaborating with people, you're publishing in journals that might not be right down your wheelhouse based on what you had done before?

MAZMANIAN: I think every few years I'm doing something that I hadn't done before. This has been the history of our lab. At this particular moment again, most of the field was being driven by sequencing technology. Very few people were asking questions in basic model systems and asking more biologically or hypothesis driven questions. It was a lot of cataloguing, a lot of just profiling of microbiomes. I would write articles describing how the field should move beyond description and more into identifying mechanism, but now I guess I've matured and I realize that something had to really jumpstart the field and it happened to be the DNA sequencing that led to the profiling.

Our perspective was a little different. We're experimentologists and we want to then take the hypotheses that people were generating in humans by profiling and then ask are there particular organisms that we see in human systems? For example, as we were doing our work in mouse models of Crohn's disease there were many other researchers profiling the microbiome of people with Crohn's disease. We knew that the bacteria that we discovered in mice was decreased or depleted in people with Crohn's disease. That in and of itself is pretty interesting because, getting back to this concept that bacteria cause infectious disease, in this particular case the absence of an organism was the risk factor for disease because that organism was protective. These are the things that we're learning from the field. Our work has been highly collaborative. It depends on the project; what type of collaborators we have and what expertise and tools and resources they bring, but in the early stages a lot of it was information, just data, from humans that we would then model in mice. Now we have collaborators for almost every aspect or all aspects of our research program. That's sort of a longer story as well.

I'll also note—and this is maybe TMI—but I find the microbiome community to be really collaborative. People who are just willing to share everything, and that's a departure from what I learned about the immunologic community. Immunologists tended to be more competitive. I know these are blanket statements and that's certainly never fully accurate, but they're generalizations. Maybe because it's just so new, but microbiome researchers just tend to be so collaborative that I always felt that it was so easy to work with others and that continues to be the case. Of course, we work with many…in fact these days we work mostly with neuroscientists and computer scientists.

ZIERLER: Would you care to guess as to why that is with micro…I mean, what do you think it is besides the newness? Is there something biologically fascinating that draws a particular kind of researcher to that field? It's all conjecture, of course.

MAZMANIAN: I don't know. I don't know why anyone would choose a particular field and what their personality or collegiality level would be. [laugh] I think it's just because it's new and there are just so many questions to ask that there's always room for everyone. I don't want to harp on immunologists, and it's not like immunology has reached a point where we know everything. We don't, but I think there tended to be just more people doing very similar experiments or asking similar questions that may have led to that competition. I've discussed with colleagues and I think there's many people who would agree that different communities just tend to have different levels of competitiveness.

ZIERLER: Sarkis, you mentioned launching the company, but chronologically we're still in your postdoc. Did you think about—

MAZMANIAN: Sorry. This was about five or six years after I started at Caltech, so I just kind of told the story very quickly.

ZIERLER: But the question there is being in Cambridge, the biotech startup community is already thriving at this point. Did that sort of influence you? Were you aware of all those things that were happening or that's really a separate story?

MAZMANIAN: The biotech community in Boston, specifically as you mentioned in Cambridge, was really just taking off when I was finishing my postdoc.

ZIERLER: OK, OK.

MAZMANIAN: I knew this because…I should be clear. I was at Harvard Medical School, so I was in Boston at the Longwood campus, but I'd go to the Cambridge campus once in a while as well. Cambridge that we know looked nothing like it did 20 years ago. It was a bunch of old warehouses. It was actually a pretty…there was a lot of crime in that area at the time. I remember just as I was leaving, maybe a few months before I left Boston, I remember seeing a large building with the word Novartis on it in Cambridge and I was wondering why did this pharmaceutical company build a site in Cambridge?

ZIERLER: [laugh]

MAZMANIAN: They were like the initial colonizers. Then I started here at Caltech in 2006 and I'd go back for seminars or conferences and then I felt like every time I went back there were like three new buildings in Cambridge. That really absolutely exploded after I left. It didn't influence me much at all. Here I had my own internal view of trying to ask basic science questions, but always do it in a way that there may be a translational component to it. What I think really inspired me to think about company building wasn't so much the health aspect, but was the entrepreneurial environment and culture here at Caltech. As you know, many faculty here have companies and I kind of see that it's happening. The Tech Transfer Office here is just really enabling. I remember being schooled by people in Tech Transfer about how you patent and start companies and fundraise my first few years here, so that when I felt like the technology was ripe enough it was never easy, but it was a bridge I was willing to cross in terms of thinking about building a company. I launched that company. It was called SymbiOX Biotherapeutics in 2011.

ZIERLER: At what point in the postdoc did you feel ready to enter the job market, that you were ready for a faculty position? When did that happen?

MAZMANIAN: It happened in the middle of 2005. That's just because…and this is in many ways how at least biologists get jobs is, you publish a paper. You show the impact. Once you have peer review I think the credibility of your discovery is stronger and that's when the major manuscript of the work that I just described was published and I felt like that was a good time for me to start thinking about jobs. The paper came out in May or June of 2005. The timing worked out well because again, at least in the biological sciences, most of the job advertisements come out in the fall. It may happen in other fields; I just don't know. I just started putting my application together in the summer and then started applying in the September/October timeframe. The timing worked out both in terms of my publication, but also when the job market started opening up that year.

ZIERLER: Just to clarify at this juncture you were all in on the academic path. You were not considering biotech, industry, that kind of thing.

MAZMANIAN: It never crossed my mind. Well, it crossed my mind, but—

ZIERLER: You made your decision.

MAZMANIAN: —I was going to do. I've never had goals in my life, David. I think I alluded to this when I was talking to you about working in a body shop.

ZIERLER: [laugh] Right.

MAZMANIAN: I've never said, "This is what I want to be when I grow up," or, "At this stage in my life this is what I want to do." I just felt like when opportunities were presented to me I felt like if I was ready and made good decisions along the way…in retrospect it didn't feel like it at the time, but I think looking back I feel like to whatever degree those doors were opening I had walked through the right ones. I just never considered…maybe I was naïve. I never considered alternate careers. I suspect I would've if it didn't work out the way I wanted it to in academia. Things were different at the time, because when I was a graduate student there was a real chasm between academia and industry. It isn't what it is now. I was talking to someone earlier today, junior faculty, telling him, "You have to be able to navigate industry to be able to really spread your wings in academia." There's so many resources and capabilities in industry, but it wasn't like that at the time. I think people who were like…maybe there was a little bit of hubris as well among academics that the best research is done in academia and not in industry. Those are the things that made me not even think about other career options. Of course, there's many things beyond just industry like law and journalism. Again, probably pretty naïve, but I was all in on the academic path.

ZIERLER: As you were surveying the market in terms of what was open, what was in demand at that point, and then on top of that really coming into your own in terms of the kind of science that you were doing, the kind of reputation you were building—where did you see yourself sliding in relative to the kinds of opportunities that were available?

MAZMANIAN: In terms of academic institutions?

ZIERLER: Exactly.

MAZMANIAN: I realized how important it is to be at a top tier university. To this day I felt like it's not a direct comparison of our technology to someone else's technology; theirs could have been just as good. But we shouldn't have been behind them if we started earlier. It's not competitive in the sense that we want to win

ZIERLER: It's the students. That's really number one.

MAZMANIAN: Absolutely. I benefit—and Caltech gives me a lot. I love Caltech. But the most valuable thing they give me is access to the most talented students in the country and the world.

ZIERLER: This is undergraduates and graduates you're speaking of.

MAZMANIAN: Of course. Absolutely. The students are the engine that drive the ship. I've been fortunate enough to train at top tier universities, so I saw among my colleagues just how valuable it was to have access to such driven and talented people. The methodology I used to apply for jobs was in that first year I was only going to apply for jobs at premier institutions. If it didn't work out I would've recalibrated and maybe thought the next year about broadening my search. The other aspect that was very important to me was geography. As I mentioned I'm from Los Angeles and it was important for me to see if I could come back to L.A. I applied to probably between eight and ten institutions only. I applied to all the Southern California universities and then a handful of universities in other places. But again, really the most premier universities. I got lucky that things worked out here at Caltech. I felt like I was very fortunate because I look back…I'm going to sidetrack here for a minute. I've been on a search committee for a couple of years; I'm actually chairing the search committee for our next hire going into fall. There's one opening for 700 applicants. It didn't have to be me. [laugh] I just feel…I don't know…I just feel so fortunate.

ZIERLER: Alright. To set the context. You're from Los Angeles, you spent so much time at UCLA, you're in biology. Before you arrive, before you applied for the job, gave your job talk, can you talk about how Caltech loomed in your mind in terms of its legacy in biology, its approach to science? Where did it occupy itself for you?

MAZMANIAN: It was like a dream. I was reading about Seymour Benzer's work as an undergraduate.

ZIERLER: Oh, wow. [laugh]

MAZMANIAN: In textbooks. There's a conclusion to this line of thinking as well. It was like this magical place. Seymour's work…because I was a microbiology molecular genetics major as an undergraduate, and he was a molecular geneticist and he had discovered many of the principles that we now know about DNA. Then reading about some of his peers like Sydney Brenner and Watson and Craig and Max Delbrück who were all heroes—exposed to all this as an undergraduate. I thought to myself, "What a mecca for science up the road." Things that were discovered there are in the textbooks I'm learning from in college. That's what Caltech always was to me. There's more of a proximal approach…is my uncle who just passed away a couple weeks ago I had mentioned, lived in Pasadena. I had driven by the institution and it just seemed so strange. I didn't understand it. I'm not sure I fully still understand it.

ZIERLER: [laugh]

MAZMANIAN: It's a very unusual…it didn't look like a university in at least the way that universities looked in my mind because it was just so small in the middle of a neighborhood. It was just more of a concept than an actual place. When I applied and was invited for an interview I thought, "This is really magical."

ZIERLER: Sarkis, I'm not sure how the timing worked out, but coming in 2006 right at the tail end of David Baltimore's presidency, the revolution of Caltech hiring a biologist to be president, and then all that he did in terms of fundraising and increasing Caltech's biology status and things like that—did that register with you at the time? Was that significant for you at all?

MAZMANIAN: It didn't register with me at the time. It has as I've learned more, as I did learn more about really the history of Caltech once I started. My view, the lens at which I was looking at it through Caltech, was in biology. I didn't understand where biology fit on the campus. I wasn't exposed to physics and engineering. Of course, we all knew about astronomy and especially seismology because we'd see Lucy Jones and other people on TV. I really learned that the heart of Caltech is that it's an engineering and physics school. My view of biology when I first started…it was still the stepchild division. I think to some degree, maybe to a larger degree, undergrads still kind of even view it that way—especially the undergrads in the harder sciences, is that biology didn't have the impact on campus and worldwide that some of these other disciplines did. But I have to say that it's been really gratifying to see how biology has grown at Caltech, but also how the life sciences have grown in academia and industry and the world. I think the tailwinds from how much biology has grown as a discipline probably fed into where biology is in its standing on campus. I still think that there's a lot we don't know about biological systems; obviously that goes without saying. The future of science, in my mind, is in biology.

A Perfect Fit at Caltech

ZIERLER: Sarkis, walk me through the job talk when you get to Caltech. Who are you interacting with? What are you speaking about from your research? How do you feel in real time that this might really happen for you?

MAZMANIAN: I was talking about the concept of curing disease with bacteria. Of course, this is all rooted in data and evidence that was building at the time, but this was the goal, the dream. Again, I was interviewing at other institutions as well. I felt like Caltech was the only place that really got it. I didn't know I was right at the time, but assuming that I was, that there was this view that you reach for the stars. If you're able to make it it's going to be revolutionary. If you're able to show what you think is happening that it's going to be a gamechanger in your particular discipline. I feel like the other job offers that I had were really because of my CV or because I had done, not because of this vision that I had.

ZIERLER: They were retrospective as opposed to forward looking in their attraction to you, is what you're saying.

MAZMANIAN: I felt like the offers were just because they felt like I was a safe bet because I had a good track record in science. I don't think they got it. I don't think that they shared this approach that you really have to think about breakthroughs and they're not going to happen incrementally. Sometimes you just have to ask the big question. And you're going to fail. We fail. We ask big questions; we fail all the time. But the ones that actually do result in validation move the field in really exponential ways. This is part of my approach now as well. I didn't realize at the time, but was always a part of the way I was thinking and I had no problem presenting that in my job talk. I presented the aspiration, the vision that I had mentioned.

I'm going to tell you the faculty I interacted with. Really Pamela Bjorkman, Ellen Rothenberg were the champions because I think they were the closest to the immunology research that I was doing. David [Baltimore] was great; I didn't have much contact with him. I'm sure he was busy. He was wrapping up. His last day as president was the day before I started. August 31st, 2006 was his last day as president; I started here September 1st. This was all happening. My interviews were happening in March of that year. I remember after the job talk Pamela and Ellen came up and there was this look in their eyes. Something I said resonated with them. [laugh] I really think it was just like they got it. They were like, "We don't know if this is real. We don't know if it's going happen, but if it does this is going to be phenomenal." I just connected.

ZIERLER: Which is, Sarkis, just to interject—that's really the opposite of at other universities where you got job offers. It was a safe bet. Your attraction at Caltech was really that it was risky to bring you on.

MAZMANIAN: That's right. It was even in the nature of the way I was asked to present both the research and the vision. We do a job talk and there's also a chalk talk where it's more unstructured, or it's supposed to be unstructured. I like to be unstructured. It is unstructured because it's not like a full presentation with slides. It's more like you and a white board and a group of people and more of a dialogue. All my other chalk talks—this happens to this day. I was asked to write one of the aims of a grant that I would submit to the NIH. I thought to myself, "I guess. That's one way to approach the problem." It just seemed so rigid. Is the objective really just to generate money or is the objective to make discoveries? Whereas Caltech was a free for all. [laugh] It was just me and people peppering me with questions and really going down these thought concepts. It was very much a back and forth in the chalk talk. I felt like it was a brainstorming session, not an interview. That gave me a glimpse into how things work at Caltech. I hope this doesn't sound immodest, but I feel like my thinking aligns pretty well with Caltech.

Two quick anecdotes. Both back to Seymour. One of the people I interviewed with was Seymour Benzer. I'm awestruck sitting there with him and then I asked him at one point, "Why did you want to meet with me?" What he did and what I studied were two different things. He said, "Well, I'm just really fascinated by…" What happened is I gave a job talk and then my interviews and chalk talk were the next day and then he got on my calendar. That's why I knew that he originally was not on my calendar. Something I said made him start thinking about the bacteria in his own intestines. He kept asking me about all the foods that people eat and how they affect their microbiome. Obviously we know a lot about that today. I think it was intuitive at the time to think that diet really shapes the microbiome. It turned out—I didn't know this—that Seymour was a foodie and he would go to these exotic places around town and maybe even around the world and eat these foods from all over the world prepared in incredibly unique ways. I think there was even a book written about this. That's why he was asking. I returned to Boston a couple days later and he sends me an email describing how he took a sample of his own feces and smeared it on a microscopic slide and looked at the bacteria in his poop and described swirly things and round thingamabobs. [laugh] He started just describing what he was seeing. I thought that was flattering and peculiar at the same time.

Then when I was made the offer to come here, Elliot Meyerowitz was the chair of biology division and he offered me space in the Broad Center which at the time was three years old or something. It was a brand new space; no one had ever occupied that lab space. There was also another lab unit on the second floor of the Church Laboratories. I toured both and it turned out that the lab in Church was right next to Seymour's lab. As foolish as I was, I told Elliot I would take the space in Broad because it was shiny and new. That night I felt just horrible because I realized that I have the opportunity to work next to Seymour Benzer. I sent Elliot an email saying, "I changed my mind. I'm going to take the lab space in Church." I was very fortunate to overlap with Seymour for only about eight months. He passed away eight months after I started. He would come by almost every day and just sort of check in on me and tell me about his ideas about how bacteria in drosophila intestines were mediating metabolic changes and digestive features of the fly. Something about our research made him start thinking about applying microbiome research to flies. He probably passed away before any of that was published. I know he didn't publish on any of that.

ZIERLER: Sarkis, as you know Caltech prides itself with its assistant professors on setting them up for success, giving them the resources they need on the basis that the expectation is the achievement of tenure because that's a win all around. I wonder if you could speak to that on a personal level.

MAZMANIAN: It's just that, David. Faculty here at the assistant level stage get the same amount of space as full professors have. We've got material resources, but it comes down to the students, the ability to recruit. Students love going into assistant professors' labs because they're most at the leading edge of their field at the time. But really equipping us with an environment with that talent that I often—I say it as a joke, but I think it's absolutely true. It's hard to fail here at Caltech as an assistant professor. I feel like most people who start here don't have a lot of stress or anxiety about getting tenure because it's the material resources, it's the students, but it's also just the support of the other faculty here, of the senior faculty. They're just so encouraging. Everyone is pulling for each other. It is hard to fail here, I think. I hope I'm not trivializing the process. I've just come to embrace that model and I have students or postdocs who go to other universities and after spending time here I tell them it's not the same in other places where they count your grant dollars and how much money you bring in dictates how much space and resources you get. Here people just believe if we have hired you—and they get to know us.

Now being on the other side I do the same thing. I really get to know the people that we want to hire. I know them obviously as people, the personalities as best I can in those interactions. We read each other's papers and really try to get to know the science to the point where there is…almost like we can talk each other's language. It's just like this perspective that when one of us succeeds we all succeed. When Caltech is recognized we're all recognized. I think that's the culture here. I've known nothing other than that. I don't see the competition that I know happens at other institutions. I think just that support for young scientists in my mind is the model because not only are we supporting them, but they're also growing where I grew up in this environment. Then you do that for the next generation. It sort of perpetuates itself.

ZIERLER: Alright. You have the vision. The faculty is excited. They bring you on board. You have the resources. What's the game plan? What's the sequence of events to putting it all together for you? What happens first? Is it students? Is it grants? Is it all of the above? What's the sequencing?

MAZMANIAN: For me it was when I arrived the lab space hadn't been fully renovated and so I spent the first month here in my office writing the grant. I had asked Caltech to at least have the office renovated so I'd be able to work there. I wrote an R00, which is a traditional NIH grant that a year and a half later got funded. That's what I did the first month. Once the lab was ready I was super lucky to be able to recruit a very experienced lab manager who helped set up the lab, meaning all of the equipment and infrastructure. Also I was very lucky that the first person in the lab was a postdoc from UCLA. She's now a full professor at the University of Utah. Her name is June Round. June applied to my lab a month and a half, two months after I started here because her husband was starting in David Baltimore's lab. They were both moving from UCLA to Caltech. June was incredible. She helped me set up the laboratory while she was setting up her project, and as soon as graduate students started coming in maybe six to eight months after that, the first crop of graduate students started coming into the lab. It wasn't just me training them, but it was me and June. It was almost like a small intellectual environment already. Graduate students started coming in, eventually more postdocs, and within—I think this is pretty typical—three to four years there were at least 10 to 12 people in the laboratory.

ZIERLER: Sarkis, in setting up the labs were there new opportunities with equipment that were exciting to you at this point?

MAZMANIAN: Not with equipment. It was more concepts; it was more sort of ideas. As you know at Caltech we do a terrific job in giving people resources in the form of equipment or even building you new things that don't exist in the world in ways that are going to just absolutely supercharge their research. Our research is not technology heavy or equipment heavy. Whatever is standard wet lab equipment is what we use, so nothing in terms of a game changer.

ZIERLER: In Caltech's famous way of having very low barriers between departments and divisions, the culture of collaboration even beyond biology, were you able to take advantage of that at all between biological engineering and physics and chemistry? Were you looking to talk widely once you joined the faculty?

MAZMANIAN: I was, and we've had periodic collaborations with chemists and with engineers on campus, but for me that collaborative environment and the fact that the divisions sort of catch all the different disciplines within that field of research. Obviously you know for people who will read this, it's not as if we have a microbiology department, an immunology department, a neuroscience department—it's all under biology. Now it's called Biology & Biological Engineering. That sets up the fact that I'm not just talking to other microbiologists all day. We're talking to people from molecular biology and developmental biology. Really what changed the trajectory of my career were conversations with neuroscientists. I'm a simple microbiologist; that's the entirety of my training. I've learned some immunology along the way, but I started forming friendships with a number of people on campus. Really the scientific friendships came with really one neuroscientist, Paul Patterson. Have we talked about Paul?

ZIERLER: Only a bit.

MAZMANIAN: This is interesting.

ZIERLER: Did you meet him early on? Was that sort of right at the beginning for you?

MAZMANIAN: It was probably within a year. I would've met people, but by the time I had sit down conversations with people obviously it was going to take some time. I'd say within a year I started having regular conversations with Paul. This is a story I've told often. Paul and I were having lunch at the Athenaeum and we had lunch because I bumped into him at the gym in Braun. We just said, "Let's get together." We really didn't know each other that well. We were sitting down for lunch and I am sure I was just going on and on about bacteria in the gut and what I thought. The crazy things that I thought they were doing. Paul was a man of few words. He passed away in 2014. I remember him distinctly. He said, "I think kids with autism have gastrointestinal symptoms. I remember reading that somewhere."

At the time it was very anecdotal, but it turned out that there were a few case studies if you sort through the literature—really not well controlled studies at all—that showed that there were fluctuating constipation, diarrhea, abdominal cramps, and bloating in kids with autism. But Paul mentioned this at our lunch and Paul was interested among other things in the role of the environment on complex behaviors. He had set up a new mouse model that reproduced many of the features of autism. He set up the model based on epidemiologic data that during pregnancy in humans if a mother has an infection she is much more at risk to have a child with an autism or schizophrenia diagnosis. There were some epidemiologic studies from Sweden and Denmark to this effect and many years later it's been reproduced. Even here in California there was a two million family cohort that reproduced this effect as well. Interestingly there appears to be a signal now with COVID that infection coupled with a high fever during pregnancy leads to neurodevelopmental diagnosis in children. Based on this epidemiologic data Paul modeled in mice essentially giving pregnant mice the flu and then showed that the offspring of those animals had behavioral symptoms consistent with autism. Mice don't get autism. Autism I think is a uniquely human trait, but they had socialization changes, vocalization changes, showed repetitive stereotypic behaviors, anxiety like behaviors which are all found in humans. Obviously the behaviors themselves are different, but they're analogous to mice.

At that lunch Paul described all of this to me, and I said to him, "I am sure you are throwing the intestines of your mice away." He said, "Of course we are. As neuroscientists we study the brain." I said, "Next time you do an experiment let me know and I'll send a student over to collect the guts of your animals and we'll just do a couple of basic techniques that we have in the lab to look to see are there changes in the intestine of mice whose mothers had the flu and mice whose mothers didn't have the flu." They were the control animals. On the first experiment we showed changes in epithelial cell biology, we showed more inflammation in the animals whose mothers had been exposed to the flu, and we showed changes in gut bacteria of those animals. I was hooked. It was literally one of those events where I didn't know it at the time, but I was making decisions that really showed that there was something there that I thought was really interesting. That collaboration continued into what I think was a really nice discovery. We co-mentored students together.

I started thinking about—this was the genesis—the gut-brain connection and how events in the gut—in our particular case how microbes in the gut—can influence activity and function and maybe even development of the brain and potentially alter behaviors in animals. As we've talked about and as you know, my lab entirely now works on the gut microbiome-brain connection. But again it was that one lunch that probably wouldn't have happened at any other university because I would've been in a building only with microbiologists spending all my time going to microbiology seminars and not having the bandwidth to really explore all these other disciplines. It could've easily been a pull towards developmental biology or something else, but the stars just aligned that it was neuroscience.

ZIERLER: Sarkis, would you say that this discovery, this new line of research was really your first big Caltech research? Or were there things before that were sort of prelude to this?

MAZMANIAN: Depends on how you describe big or define big. I felt like we were doing pretty good work, but it was really along those same lines that I described that were an offshoot of my postdoc of thinking about microbe-immune interactions and microbe-microbe interactions. I think there was just like a sea change if you will. To go into an orthogonal direction and think about the brain and think about neuroscience and behaviors. I feel like we had made some nice discoveries prior to that, and we continue to, because as we sun-setted all the previous projects that took many years as well, and as we ramped up our work on the gut-brain connection. I think this is one of those things…and Ellen [Rothenberg] mentioned something to me during my job interview that was a harbinger to this or predicted this. She said something to me that I've never forgotten. She said, "You can go to another university and maybe someday you might be a great microbiologist or a great immunologist, but if you come to Caltech someday you might be a great scientist."

ZIERLER: Yeah. Ahhh, that's an awesome quote. [laugh]

MAZMANIAN: And, you know, she was right. Not about the great part, but the fact that your career goes in a really unique trajectory that you can't define early on.

Connecting the Brain and the Microbiome

ZIERLER: On that point, to go back to your job talk and this vision of discovery. What aspects of that original vision do you see as the building blocks to this gut-brain research that happened later on as prelude and what aspects were totally new because of this random encounter and the random lunch and going in a direction that probably might not have happened except for these serendipitous circumstances?

MAZMANIAN: I think the foundation that's almost been like a unifying principle—thought it's a loose one—was the fact that instead of looking at the cataloguing or studying the effects of microbes I've always thought to myself that unless you have a molecule, like a fine structure, you really can't unravel very deeply the biological underpinnings of an outcome of a phenotype that you're studying. Don't get me wrong. You think about astronomy or even systems neuroscience where a lot of it is very observational, but you can't get down to any sort of handle, like a molecular cellular handle, because of just the nature of the way you're looking at the problem. But that opens up others to come in and now explore things on a more, sort of granular level.

I've always approached things more from the bottom up if you will, where it was very important to identify particular structured molecules that are mediating an event because then you can really get down to cell types that respond and receptors that bind and how it changes the activity or the function of a cell or tissue or an organism. Again, the molecule is the trigger; it's the light switch that we're mentioning. That part has stayed consistent to this day. When we see a change now it's behavior or neurodegeneration what have you and we know that based on the microbiome of the animal that's making that change I always think about what are the set of experiments that we need to do to get down to what are the molecule or molecules that are mediating that effect? That opens up all sorts of new avenues for research. What was different as we mentioned, is this fact that this concept didn't just apply to the immune system, but it applies to the nervous system and the metabolic system and the endocrine system as well. I guess the breadth of biology that in one way or another is tied to the function of the microbiome.

I'll zoom out a little bit. There are many people who would view the microbiome as an organ, as its own organ system. It just happens to be an alien organ system to us, but it has functions and interactions with all components of the body. I think this is more what the effects on the brain taught me is that not to think about the effects of the microbiome just on the immune system or on the gut itself, but really think about how that connects to all the aspects of the body. If I can just then bring it back to that university analogy and I think this is one of the things that really works well at Caltech. Many other institutions and medical schools and medical centers have departments of different disciplines—microbiology, neuroscience, immunology, what have you—but that's not how the human body or any body or any organism works. Everything is talking to everything else at the same time. I think when you silo different disciplines it becomes hard to train graduate students to think about the all the different interactions that happen in a complex organism. It's not intuitive to be able to learn how to study at that level of complexity. I think one of the ways that the structure at Caltech really help us do this is that we are in our own disciplines exposed to so many other disciplines. It makes that leap to studying these interactions between organ systems much easier or the activation barrier to actually being able to study that much lower because we're exposed. As a microbiologist with some immunology experience I was exposed to more neuroscience than I would have been anywhere else. I really believe that's because of the way the departments and divisions are structured here at Caltech.

ZIERLER: Sarkis, last question for today, and it's one that touches both on your style as a mentor and a collaborator with your peers. As you become more involved in neurological disorders, where is it important for you to read up for yourself and where are you outsourcing that to people who by just dint of their own research expertise are far more knowledgeable about those kinds of topics? How does that work for you?

MAZMANIAN: It's a dynamic process because I feel like I'm always learning. I feel like I have so much to learn about neuroscience, and in fact I would not ever consider myself a neuroscientist. At least not yet, maybe a decade away from that. I'm probably like year one in an undergrad neuroscience program still. I know the basics. I've just learned the basics over the years, but there's so much nuance because it's such a mature field and there are so many technologies that apply to the field that I feel like I'm always playing catchup. The outsourcing goes to collaborators who bring in not just general neuroscience knowledge, but domain-specific knowledge, a lot of it to the students and the postdocs in the lab. They educate me. When we meet and talk about their projects I'm learning basics, not just what's in the project, but basics that have already been in the literature maybe in some cases for decades. I learn this through organic discussions as well. In addition to the reading that I do and the seminars and meetings that I go to it's those conversations with students that really help me sort of catch up or speed up my education.

ZIERLER: As the lab is reinventing itself, how do you manage that in terms of shepherding through graduate students who entered when you were working in one area and now you've moved on to another, but that's not what the graduate student was there for? How do you manage that with students and the ones coming in to make sure that they all get the best of your attention and the best possible mentorship from you?

MAZMANIAN: This happened several years ago as we were sun-setting the previous projects, where people who worked on those projects continued to work on their projects, but it was really the ramp up of the people coming in that started working on the gut-brain connection. I never sort of announced it, but when I was having conversations with incoming students or postdocs I let them know that this is where I felt that there was an opportunity. This is where I felt like there was a growth area and our ability to make an impact in a relatively new area of research. I would describe my enthusiasm to them and I think that's how they learned or they understood where I thought the frontier was at least for our research. I've always been very upfront with students in two respects. I tell them this is where I'd like the laboratory to go, but also in the fact that they design their own project. I rarely if ever have told someone that, "This is something you have to work on because either I'm interested in it or we have funding for it." I want students and postdocs to follow their interests. It's a collaboration. If they see what I'm excited about then they can maybe develop their own interests within that particular area, but it's never dictated to them that this is what I'm interested in and so this is what a person must do. It's an agreement.

Oftentimes I have students who don't want to do the things that I'm most interested in. They develop projects that are outside of that and I try and give them as much attention as I do everyone else, but I think naturally we do have our own interests, and students know this. As I'm upfront with them I'll tell them my time, my bandwidth is limited. The more independent they are—I think this is a win-win, right? The more independent they are, the more likely that we'll be successful in our laboratory because they have to take ownership of their projects because it's something that they're interested in, something that they discovered. The win-win part is because by them being more independent, I think it just prepares them for the next stages of their careers. It's not as if I or someone else in the laboratory was helping them. They had to do all the basics to build their own projects, and I think they require a particular skillset in doing that, where that allows them to then replicate that when they're independent. It's not as if people come to our lab and pick up where someone else left off. They come to our lab and say, "What am I interested in?" It's almost like going back to the drawing board every time.

ZIERLER: Sarkis, this has been another terrific conversation. In our next one I look forward to the origin story of the company and what happens next.

[End of Recording]

ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Friday, August 12th, 2022. I am delighted to be back with Professor Sarkis Mazmanian. Sarkis, once again great to be with you. Thank you so much.

MAZMANIAN: Hi, David.

ZIERLER: What I want to do today, Sarkis, I first want to get the chronology and the names correct. We're going to pick up right from our last discussion—your entrée into the startup world and thinking about translating the science into businesses. I want to make sure I get this right. In 2018, it's Symbiotix Biotherapies and then in 2019 it's Axial Biotherapeutics. Is that right?

MAZMANIAN: No, there's three companies total now. In 2011 it was Symbiotix Biotherapies. 2016, Axial Therapeutics. Then in 2020 I started another company which I forget if I've mentioned this to you. It's called Nuanced Health.

ZIERLER: I was looking at press releases, so there must have been something in 2018 and 2019 that was significant.

MAZMANIAN: There's always ongoing things, but the inception of the companies were those dates.

ZIERLER: Alright. Now I want to bring it back to—this is something that we covered a little bit in our very first discussion. That was Caltech's institutional culture and response to just the idea of thinking about bringing the science into a business setting. Before you even talked to anybody on campus—colleagues, administration, people like that—what was your sense of how these ideas that you were germinating would be received circa a couple years before 2011…2009, 2010?

MAZMANIAN: Received on campus?

ZIERLER: Right.

MAZMANIAN: I'm not sure. I can't recall giving that question any thought in terms of if I were to start a company how would that be viewed by colleagues or the administration. In fact, if anything—and this speaks to Caltech opinion—it was like an iterative process before I realized…once the process started I realized what Caltech's role in that would be. So, I didn't give it thought prior to that. I forget the details of what we've talked about, but I may have mentioned Caltech, especially the Office of Tech Transfer, and a couple of colleagues in BBE were incredible in terms of their generosity with their time guiding me through even the basics. I'm happy to go into details and names of people. Even the basics of how companies are started, but there was overall a sense of really a lot of encouragement from the administration and from colleagues who had gone down this road before.

ZIERLER: Maybe part of it if you didn't think too hard about it, it's that this was really not a big deal at this point in Caltech's history.

MAZMANIAN: I knew that many faculty at Caltech started companies, so maybe that's why I didn't think about how it would be viewed is because I felt like there were paths. Maybe not one, but multiple paths to doing this and that it wouldn't be viewed negatively in any way. I know there are certain academic groups, I guess not really institutions, but HMI and others…at least outwardly they are more limited or sort of restrained in their enthusiasm for starting companies because they want people to focus on science. I get that; I don't disagree with it. I just think that starting companies just exposes you to an entirely new world that helps both basic research, but also the ability to help people and translate that as well. I never worried that Caltech would be opposed to or resistant to these ideas. There's really an external set of activities outside of campus that motivated me to start the company.

ZIERLER: Among those colleagues in BBE, whether it's specific individuals that you want to talk about or experience, what were some of the through lines in the conversations you were having about how to get this started on the right track?

MAZMANIAN: Two people really helped me. Mostly it was Ray Deshaies. Ray was a faculty member here. I don't know if you know of Ray. He's now senior vice president of research and development at Amgen, but he is a National Academy member, a Howard Hughes investigator here for well over 20 years. He was assigned as my faculty mentor. That's what we do in BBE is that when you start as an assistant professor the chair essentially pairs you up with a senior faculty member and then you just go have coffee or lunch every so often. Nothing structured. Ray and I would meet pretty regularly and he had infinite knowledge and wisdom and was very direct about sharing it. That's just who Ray is. He'd ask me, "How's the lab going?" I said, "Fine." We immediately started talking about the company because he was just so excited about entrepreneurism and then he went off to lead a biopharma company. He would give me such detailed advice. I didn't know where to start on so many different parameters in terms of what are the steps, how do you deal with relationships with investors, what should I expect personally both in terms of my time commitment and my effort and my compensation? What are the lessons he's learned and the mistakes he made? Just a broad spectrum of advice. That was super helpful. This was literally, David, almost like every week I'd chat with him about something.

The other person who really helped me a lot was Steve Mayo. This was before Steve was chair. Steve had started at least two companies that I know of, one of which is now a publicly traded company, a very large company. He was super helpful both in terms of giving me advice on the minute-by-minute timescale that Ray did, but also even introduced me to investors and to venture capitalists to help fund those companies. They were really…again those two really stood out, but I talked to a number of other people casually as well. Those conversations reinforced the prior question that there is a lot of enthusiasm, almost organic enthusiasm here for starting companies.

ZIERLER: Sarkis, this false binary between science and business that venturing into that world would take you out of the science—did you share that yourself? In other words, were you sort of part of the Academy in the assumption that those were separate things, but that what you were doing at Caltech pulled you into business considerations that you might not naturally have gone into yourself?

MAZMANIAN: When I was a graduate student I was trained to think that Pharma was the dark side.

ZIERLER: [laugh]

MAZMANIAN: You don't deal with those people for two reasons. First, they're just trying to steal your ideas.

ZIERLER: Right.

MAZMANIAN: Right. This is just what I was told as a naïve, impressionable graduate student. Also that the smartest, most talented people wouldn't go into industry. They're in academia. Why would you even want to deal with them? Whether or not that was true 25 years ago I can't say, and obviously this is in the eye of the beholder. It depends on which environment you're in and so forth. It's sort of not a universal truth, but I think there was more truth to it than there is now. What has happened in the last 25 years that has I think given academics a lot more reason to be accepting of industry is that industry has almost across the board sun-setted a lot of their internal R&D and now relies on academia and biotech for the innovation. Then they come in and they just essentially develop the assets once they've matured. They're not identifying new targets like new molecules, new receptors, new pathways to drugs and so on and so forth—at least generally in a way they were doing a few decades ago. That's helped build relationships between industry and academia.

Going back to what we talked about earlier is that I learned—again, there was a set of activities that led me to thinking about starting the company. Then I started realizing just how enabling that would be. Just the ability to talk about your ideas with people who are thinking about the problem from a different way is helpful in general. They bring resources, they know how to commercialize or develop ideas that we develop in the laboratory, but also in terms of thinking about what are the most relevant activities to the development of a technology? As academics, I think a lot of times it's easy for us to do things because they're interesting or cool or exciting or novel, but those activities may not directly translate. That's fun; we should be doing those anyway. Those activities don't necessarily translate into developing or pushing the technology forward. Oftentimes those are just intellectual curiosities that we may be pursuing.

There's a certain path that's almost definable at least in therapeutics, which is the world we play in of how you develop drugs. If I know a technology, a drug that we're working on in the laboratory…for example, if I know what's going to be required to get that into the clinic in three or four years now, then I'm much more efficient in the activities I'm performing because I know what's going to move technology forward and what is not, is sort of neutral. You do the discovery and the intellectual aspect as well, but now what will be layered on which we didn't have before are those activities that I know are part of the drug development campaign. In the end that's been super helpful in making our academic work more efficient and ready to translate.

ZIERLER: To be clear it was the science—what you were seeing in the laboratory—that started you thinking about possible business opportunities and not the other way around?

MAZMANIAN: The story is this, David. I was approached by a venture firm who had read some of our papers. I can't say I'd given much thought to starting a company. They said that they wanted to build a company around this technology. I'd done some of the work here and some of the work previously at Harvard when I was a postdoc. My postdoctoral advisor and I were the people who would have been the scientific founders of the company. It's a huge venture fund that we were talking to and it was almost flattering that this entity had recognized our science and thought it was worthwhile and was going to put money and build a company. Over six months of conversation I just realized that they were really just trying to take our ideas. This became almost obvious when—I could write a book about this experience to be honest with you—people at the venture fund who were hired to start this company told me that the venture fund was writing patents behind our backs based on the conversations we were having. They were basically sucking the ideas out of mine and Dennis's heads in these conversations and really had no genuine intention of building a company around us or the technology.

ZIERLER: Did you think about lawyering up at this point?

MAZMANIAN: I did not. I may have thought about it, but I did not because what we did was we just walked away. I think I had the strength to walk away from that opportunity because I realized I saw what was happening. Sure enough as soon as I walked away and Dennis walked away—Dennis, my postdoctoral advisor—the rest of the team that they had hired to work on that company, they were let go by the venture fund. Those who came back to us said, "Why don't we just start our own company without the shadow of what this VC was trying to do?"

ZIERLER: So, that was definitely the right move, walking away. That accomplished everything you needed it to.

MAZMANIAN: Absolutely. To the point where I've run into these people again and have brought this up outright with them. I realized that this is just how that fund operated. I would say it's very atypical. Maybe not very, but it's atypical. That planted the seed if you will. I thought to myself maybe there is something worthwhile here. Maybe there is technology here that was valuable or would be useful, and I started to think about how do you start companies? That's right when—again I was already talking to Ray about all these things, but that's when I think Ray really helped. The other person who helped—I think he might be at JPL, but is no longer here—was Dan Broderick. Dan was in the Tech Transfer Office. He might've been associate director. I know he moved to JPL and was handling JPL patents. Actually I don't know if he's still at Caltech; he might be. He would just sit there for hours and just walk me through the process of, How do you incorporate a company? What is a cap table—a capitalization table? How do you value a company? Once you bring in money, how does the value of that company change? These basics, things you would learn in business school—maybe like the first week of business school. [laugh]

ZIERLER: [laugh]

MAZMANIAN: The basics. Again, out of the kindness of his heart he would just spend hours with me just teaching me and eventually Dennis and I were able to go out, raise external money, start a company. That company raised $40 million—it doesn't exist anymore—and actually got pretty far. It was sort of a circuitous route, but that was really how it started. This external, dark force that got me thinking about building a company. I'll just say maybe in addition I made a lot of mistakes with that company, but I learned a lot. I think that's helped me with the two companies that I'm working on now.

From Lab to Venture

ZIERLER: Let's get now specifically to the science. What were some of the advances that made what you were doing in the lab commercially viable even as a concept?

MAZMANIAN: First is the concept that bacteria can be potential drugs, that they have benefits. In fact, I referred to this I think when we were talking a few sessions ago about why I was hired at Caltech is that people here thought, "Wow. We always think of bacteria as these harmful, insidious, little creatures, but what Sarkis is telling us is that no, there's good ones out there." They may actually be conferring health, that sort of concept. When I was interviewing in early 2006 that was very much the concept. By 2009, 2010, there were other people thinking about this now as well. People were much more open to the idea that someday there may be a pill with live bacteria in there that could be a drug. Really I think this comes down to the actual demonstration that indeed there are bacteria with identifiable, measurable benefits. There are a number of ways to satisfy that criteria, and for us it was developing animal models of disease. Initially we started with models of Crohn's disease and models of multiple sclerosis which are essentially autoimmune diseases or forms of autoimmune diseases. And we showed that by these mice that were destined to get sick—either inflammation in their intestines or inflammation of their brains—that we were able to give them a bacteria that we identified in humans, that we cultured from humans, all the parameters associated with those disease models improved.

I think when we're able to show data like that, which the first papers came out in 2008, 2009 was really eye opening for people because it went beyond a concept. Using models that drug companies were employing for drug development we were able to show that a live bacteria was able, in some cases, to outperform drugs that were being given to people at that time. It's hard to argue against that type of data; whether or not that would translate to humans at this point is still unknown. Using those existing models to show the benefits of live bacteria I think was pretty powerful. Then we went on further. I'm making this too detailed, but we identified not just the bacteria, but the specific molecule from the bacteria. That gives a lot of comfort to people because now it's like a chemically definable entity, not just this amorphous bacteria doing thousands of things. Here's this one molecule from the bacteria that we've isolated that has the same beneficial effect as the entire bacteria itself, identifying which immune cells the molecule was interacting with, which receptors and so forth. When you have a deep mechanism of action, people are more likely to believe that indeed you're studying a biologically relevant process and not just a phenomenon.

ZIERLER: The idea that you were bringing new ideas to Caltech which is so forward thinking even in and of itself, what about the regulatory framework when you had to start thinking about FDA and other agencies, and clearly this was not on their radar? When did the company get to a point when you would start thinking along those lines?

MAZMANIAN: The initial Symbiotix company got to the point of speaking with regulators four or five years after it was started. That happened for the business reasons I can't comment on because what I now believe—and again, every business is different—is you have to engage regulators as soon as possible. I need to know what they want to see so that I can perform the activities to deliver to them. At the time the FDA was also trying to figure all of this out. They had no guidance on what are now called LBPs, live biotherapeutic products. As you may know the FDA, the way they operate is whatever the drug category class is, there's guidance and they publish this on their website. What do you need to do from a regulatory standpoint to get an approval for that modality? Think of an antibody or a small molecule cancer drug, for example. These have different guidance. There was no guidance for a live bacteria. The FDA was really trying to figure this out and to their credit what they did is they invited me and other people like me to the FDA pretty regularly. There was period where I was going out to the FDA two to three times a year to both educate them and learn from them on what was their thinking and what were their guidelines they were going to start putting together around this type of technology. It was. I would say, more of a partnership learning and educating than me just trying to understand because there was no framework for that at the FDA at the time.

ZIERLER: What were you learning from the FDA given that you emphasized this was a two way street?

MAZMANIAN: They were so open minded to the concept that they were quite bullish about live biologics, because, in the way it was articulated to me, the FDA is concerned about safety. Safety and tolerability. They're not there to figure out if a drug works or not. That's not their mandate. They felt so little risk in a drug that came from humans, that had been in humans for millions of years.

ZIERLER: [laugh]

MAZMANIAN: Effectively the safety tolerability toxicology studies had already been done by society, by civilization. [laugh]

ZIERLER: By evolution.

MAZMANIAN: Right. These organisms had evolved to not…after thinking about this, to not cause disease. That wasn't their…it's hard to make evolutionary arguments of what the impetus for any activity is, but in fact they wanted to confer health because the healthier their host was—we—they had a more nourishing environment to live in. I think to this day the FDA remains super bullish about technologies where essentially you're taking a drug from one human and putting it into another. They see very little safety issues. In fact, the data bear that out. Thousands and thousands of people have gotten fecal transplants from human donors. Whether it works or not, that's obviously a long discussion and without a clear answer yet. It's a case by case issue, but there are clearly no safety concerns. To this day the FDA will very happily give out an IND designation, an Investigational New Drug designation, for an organism that comes from humans. That thinking was even going back over a decade.

ZIERLER: Dating the beginning of Symbiotix to 2011—what exactly happens there? What sticks out in your memory as, that's the year?

MAZMANIAN: That's the year because that's when the incorporation documents were filed in Delaware. That started the activities where we were able to recruit a CEO, we were able to go out and raise money, and move the technology along. The date comes based on the actual filing dates.

ZIERLER: Tell me about the infrastructure—setting it up, getting space, finding people. How did all of that play out?

MAZMANIAN: For Symbiotix, it was a very lean company, meaning that we used a model that many other companies use for outsourcing all of the activities. Symbiotix never had labs. It was just a team of executives and all of the research was done either in my lab through sponsored research and Dennis's lab at Harvard or through CROs, commercial research organizations—essentially fee for service organizations. Not much infrastructure to build. I can guide you through the process of Axial which is much more traditional. It's a brick and mortar company. If you want to maybe fast forward to 2016 I can walk you through that process which I think is what you're looking for.

ZIERLER: Let's wait on that because I want to hear about the mistakes you made early on which clearly informed what happens in 2016. Was it a mistake to keep the company as lean as it was to start?

MAZMANIAN: I think embedded in the answer to that question in my mistake was that I didn't identify the best possible leadership for the company. The best possible leadership sets that strategy of whether you're lean or not and overall, how the company is viewed. In some ways I probably didn't have the confidence and I didn't have the patience to wait for the ideal CEO or business leader, whatever the title may have been for the company. I think in many ways, and I won't go into names, but this may be sort of personal. It matters so much because at the end of the day I do have a day job. There's only so much that I could contribute in terms of bandwidth, but also it's not my experience. It's not what I was trained to do, to start companies. I should have realized what an oversized role the CEO has in the company, and I should have made sure that I was patient and I'd vetted people until I found the right person who both had maybe the business experience or the knowledge that I was looking for, but also really the conviction, the belief in what we were doing. If you don't have that, then on a day-to-day basis you're probably not going to be driven. I think that was the first mistake, just not being patient and not going through…what I should have probably done is an organized process of identifying the CEO.

Coupled to that which we may get into with the Axial discussion, is identifying investors who were partners, or just looking to triple or quadruple their money, but who really believed in me and what I was doing, the company, the technology. Because these roads are never smooth. [laugh] There are always bumps and you want to make sure that you have people working shoulder to shoulder with you when those bumps happen, and they do. Because I was a first time entrepreneur I didn't have the confidence—and maybe that's reasonable—to wait for not just money, but smart money from humans, people who I would enjoy working with. Symbiotix existed for I think eight years, and in many ways, I feel like we overachieved based on the fact that…this is all forensic. I'm sort of looking back now and trying to put the pieces together. Based on the lack of the most ideal leadership and investors, I'm actually proud that a company that failed took so long to fail.

ZIERLER: [laugh] That's great. Sarkis, let's say you didn't have the emotional intelligence to realize that the CEO could not have been you. You mentioned your day job. Just in terms of how many hours there are in a day and expectations from Caltech, could you have been the CEO? Is that something that's within the realm of possibility?

MAZMANIAN: It is. It probably would have been a huge mistake for me to be CEO of my first company unless I had support. Unless I had mentors who would guide me through the process. I think I learned so much in being involved that I'm probably in a better position now to do that, which I did contemplate for Nuanced Health which I started in late 2020. But…I feel like…I don't know. I feel like I'm a scientist. I'm trying to be honest with you. I see both the CEOs that I work with now who I'm great friends with. I see their lives and I don't want their job. I just don't. I honestly don't feel like it speaks to my skillset because it moves…I think I'm good at certain things. I know I'm not good at many things, but it moves me from being creative to just being a manager. I'm not a good manager. I don't know if we talked about this. I let people do what they want. I think I'm a good manager when it comes to academia.

ZIERLER: You're a good mentor when it comes to graduate students and postdocs.

MAZMANIAN: Yes. I think there is a distinction between leadership and management. I think the role of the CEO is very organizational, very administrative. Again, that's just not me. It takes me away from everything, my strengths. I can only say this now after about 10 years of being in this world. I wouldn't consider it again. I did consider it a couple of years ago. I was actually thinking about this a day or two ago. I would have a nervous breakdown if our lab wasn't being as productive as I'd like it to be. It comes down to hours in the day. It comes down to where my focus is. If the lab is not productive, even if I'm doing all sorts of great things outside the lab, I would be really quite miserable. I've realized that about me. It's the lab. I don't want to trivialize this, but Lebron James told another basketball player once, he said, "Keep the main thing the main thing." For me the main thing is being a professor at Caltech.

ZIERLER: There's also the moral dimension of the disservice to your students and postdocs if the lab does not have your full attention.

MAZMANIAN: It's all part of it is that… if I'm not being a good mentor that's going to eat at me. If we're not being productive because I'm not being a good mentor because the students are not trained and they're not getting the input from me that they should be—that would eat at me. It's personal, but it's also a reflection on the fact that I wouldn't be doing my job which is to train students which I take a lot of pride in.

ZIERLER: Obviously you didn't need any external encouragement from Caltech to keep your eye on the ball, but I'm just curious in talking with the division director, provost, whoever it might be. What have you learned about the expectations, spoken or unspoken, that when a professor goes down this road that's fine but those balances need to be maintained?

MAZMANIAN: I think there are structured ways of addressing this. Many faculty will take sabbaticals. There's IP reasons. There's legal reasons to do this as well because if I'm inventing it's hard to know if I'm inventing here versus inventing at the company because: When did the idea go through my head? [laugh] Was I driving to work or driving to a meeting for the company? So, they put up these firewalls, but also to obviously free up your time to do that as well. I think a lot of this is the way Caltech operates. No one is really telling you what your role is or what your responsibility is. We're trusted to find that balance ourselves and if I want to spend more time in outside activities I feel like I have the freedom to do that as long as I'm teaching my classes, funding the lab, graduating my students. No one is looking over our shoulders. I wouldn't say it's encouraged, but I wouldn't say it's discouraged either. I think it's more along the lines of just what I said. There's just an accepting attitude here, throughout Caltech. If you want to spend your time in this phase of your career or more so in the company there's really no expectation one way or another.

ZIERLER: Sarkis, now let's go back to a very interesting comment you made earlier. The learning process where you realize that the venture can actually improve the research; the fundamental science can be enhanced by taking a business lens approach to it. Tell me intellectually how you came to that appreciation after the company got going.

MAZMANIAN: The details would be that we have a data package that we generate in the laboratory for Symbiotix. I did the experiments that I thought were interesting that were going to teach me something. I go show that data to venture capitalists, to pharma companies, or to key opinion leaders in the business world. They said, "Well, that's great, but there's just a critical piece of information missing. I cannot evaluate your technology." Really simple things. As academics in biology, oftentimes we'll identify a certain dose or a concentration of a drug or a chemical, whatever it may be that works. Then you just use that concentration to generate your data, to collect your data. Then you go and take this to a drug developer and they say, "Well, I need to see a dose response card. I need to know what is the minimum concentration where it doesn't work, the maximum concentration where it leads to toxicity or side effects, and what is the optimal dose there?" This is like drug development 101, but not something I ever thought about until people looked at the data we were generating and said, "There are critical pieces of data missing." I learned that through these discussions, through taking the data and trying to commercialize it and people telling me, "I can't help you here. I can't advance the technology because I don't have enough information. You didn't do all the right experiments."

ZIERLER: Sarkis, when did you start to figure out that the first CEO you had might not be the best fit and how did you go about changing course at that point?

MAZMANIAN: There's no one activity. I think a lot of it was based on after a few years of instinct telling me that we should be doing better than we were.

ZIERLER: What are the metrics? What does better mean in this context?

MAZMANIAN: In this context it was…so Symbiotix was one of the first biotech companies started in the microbiome space. Four or five years later many other companies had started after us and they were much further along than we were. I'm thinking to myself everything isn't apples to apples, but I felt like we had enough solid science here that we should be moving further. We're being leapfrogged by other companies. I'm thinking to myself, maybe it's not fully a science issue; maybe there's a business issue of why we weren't further along.

Then, just sort of observing the CEO and just after a while not realizing or appreciating that maybe people across the table didn't have the confidence in him that I would've liked them to have. You're not going to go into business with someone unless you believe that they can hold up their end of the bargain. I don't think that potential partners we were talking to had the confidence in the CEO. I saw that dynamic and I think that's what made me realize those shortcomings.

It lead to, for a long time, just coaching and trying as a novice myself, but someone who had lived 45 years or so at the time, to guide him through or coach him through some of the personal aspects of it. Just to get more granular: don't overstate the data. If the data are telling us something don't say it's much more than it is. People see right through that. And deliver what you promise. It's basic things, but those are some of the areas where I felt there were some deficiencies. It was very much coaching. By the time Dennis and I had gotten to the point where at least I was very confident, we needed to make a replacement. I think the company was already on its last legs and didn't survive much longer after that.

ZIERLER: Sarkis, given the idea that there were other companies doing this, did you feel like you were in a race? Did you use that to positive benefit at all?

MAZMANIAN: I never felt it was competitive in terms of time, in terms of a race. I just felt like our technology was solid. I couldn't rationalize—and to this day I do and I'll resurrect that technology someday. To this day I felt like it's not a direct comparison of our technology to someone else's technology; theirs could have been just as good. But we shouldn't be behind them if we started earlier. It is competitive, but that's more metrics. It's not competitive in the sense that we want to win. It's more like a metric or a way to measure where we were as a company. You need a reference point, I guess, to know if you're succeeding or not. The other companies were that reference point.

ZIERLER: Going back to 2011, what was the timescale that you were envisioning or that people with experience in this space suggested that you should be thinking of—defining that however you would define success—bringing a drug to market, having your investors get paid back? What are the timescales, what are the definitions you're using for success going back to 2011?

MAZMANIAN: Hard to know what those definable metrics were, like after two years we should be here or after three years we should be here. It was more comparative. Both intuitive looking back saying we could've moved faster, we could've been more efficient, but also benchmarking against other companies, yet knowing the science underlying those companies where they were and why is their drug in the clinic three years after they started and we started five years ago and we don't even have deep discussions with the FDA yet? I think after a while it started becoming quite obvious that we were lagging behind. It was a number of different events, but pretty, pretty clear actually.

The Vital Importance of Good Management

ZIERLER: Now…Monday morning quarterbacking here. You said that by the time you decided to change leadership the company was already on its last legs. Had you shifted earlier on do you think the company would have been viable even today?

MAZMANIAN: I do, I do. I can't know that for sure. I do because I think the fact that we were able to raise between $30-40 million there's maybe a lot of external validation in it. I'm trying to disconnect it from the other thing which I believe is still true, which is that I think the technology is quite important and really can help people. I see the data. It's not just our data. I feel like I can emotionally distance myself from data we've produced in the laboratory. It's not just that other labs around the world have reproduced and extended our data, but the company was able to reproduce and extend the data. I felt like the foundation was quite solid. There's generally two things. There's many things to a company. It's the science and then leadership. I'm going to get vulgar here. Sorry. I think a good management team can turn shit into gold and a bad management team will turn gold into shit. [laugh]

ZIERLER: Yeah.

MAZMANIAN: We work on shit. We work on bacteria that live in shit.

ZIERLER: It's perfect. [laugh]

MAZMANIAN: That's what I really believe. At the end of the day like everything else, David, is no matter how good the science is you have to gift wrap it, put a bowtie on top, and you have to communicate that to other people because the name of the game is getting other people excited about the things that you're seeing. I really think it was the business part. The packaging of the business and including the activities, but the packaging of the business is where Symbiotix fell short.

ZIERLER: We talked in our very first conversation about those all-important red lines to establish between your students on campus and workers at the company. Was Symbiotix around long enough where you saw it as a career opportunity after students had passed through your lab?

MAZMANIAN: Because it was so lean those roles weren't there at Symbiotix. Symbiotix at its height was three employees. But Axial is. Axial which is hopefully not at its height, but looking back the headcount is the highest it's ever been; it's 27 people now. There are a number of folks from our lab. They've never been full-time employees at Axial, but they've done fellowships or internships at the company. It has been a vehicle for those students in the lab who are entrepreneurial to sort of dip their toe in the biotech pool if you will. We had someone who was an undergrad in our lab for two years last year transition to Nuanced Health as a full-time employee. It is opening up those possibilities. I encourage that as well with all the proper firewalls. For example, we had a graduate student who just defended in June. She spent three months at Axial. It had to be remote because of the pandemic, but she spent three months at Axial. She took a leave of absence from Caltech, did an internship at Axial, and then came back to continue her graduate work largely for IP reasons. I do think it's helpful for students in the lab. Being at Caltech, many of our students are entrepreneurial.

ZIERLER: Sarkis, you mentioned your intention to resurrect this technology at some point in the future. Between the fact that there are now other companies operating in this space, how can you ensure that what you had done remains yours both as a matter of intellectual property, and scientifically so that whenever you are ready to get back into it somebody didn't beat you to the game?

MAZMANIAN: In terms of other drugs in the space we can't slow that down. To the best of my ability to evaluate this none of the newer technologies have advanced the science beyond where we had taken the Symbiotix science or the science from our lab that was in Symbiotix. It really is the intellectual property, the patents. In fact, one of the good things that Symbiotix did was build a very strong IP fence around this technology. We were one of the first companies, probably one of the first entities, to be able to get composition of matter on a natural product. What that means is a molecule that's found in nature, we were able to own the structure of that molecule—which if you read in the patent textbooks this is supposed to be impossible. If I were to synthesize a chemical in the lab I can own that structure. If it's found in nature no one could own it the way the patent office had operated. Symbiotix spent two million dollars in two years on an IP strategy that ultimately worked—again this may be granular, but may be interesting to you.

The molecule that had the biological activity that I was mentioning to you is found on the surface of the bacteria. There's the bacterial surface; the molecule just kind of sticks out right there and then it does all these wonderful things to the immune system that cure disease in mice. Once we took the molecule off of the bacteria we made the argument that that molecule separated from the rest of bacteria doesn't exist in nature that way. We own the structure that is now uncoupled from the bacterial cell, so we can now own that structure because in nature it's only found chemically bound to the bacteria. That is what's going to protect the technology. The double edged sword is that there's a clock ticking. Patents have particular expiration dates and investors know this and drug companies know this. The longer we go from the filing of that patent which was in 2017, the less valuable this technology becomes. If you can't own a molecule, you can't exclusively profit from that molecule. Investors and drug companies are less excited about that obviously.

ZIERLER: Sarkis, last question for today and it will serve as a cliffhanger for our next talk. How far along was Symbiotix and your own experience when the idea not just specifically scientifically for Axial came along, but where you were now on a learning curve where you saw an opportunity to start a new business that could build on all that you had learned from the first time around?

MAZMANIAN: If I hadn't started the company based on the learnings from Symbiotix was I ready to start a company or what was the process for the second company?

ZIERLER: Even before the science became apparent to you for a new company. You're always trapped to a certain degree within the first company that you started. What were the aspects of it that you thought to yourself, "I've now learned enough where if there's an opportunity to start a new company I have all of this behind me now where I can do it better second time going."

MAZMANIAN: I'm not sure I'm still there yet in terms of believing that I know all the things I should know. Maybe I'll never get there. It was a confluence of emerging technology in the laboratory based on the basic research we were doing that I felt was valuable, but it was also—and I remember this distinctly—me starting to realize all of the things we've been talking about in the past few minutes that Symbiotix is not living up to its potential. It's not as if, from my background in Symbiotix…I worked in Symbiotix many years after I started Axial. I felt like we had something new. I had learned a few things—not everything. I just didn't have the faith in Symbiotix after—this would have been five years into a company. It was really those things.

Unrelated to what you're asking externally, but this is I think important if we want to dive into this next time we can. The second time around was much easier. Not just from what I had learned which I think I incorporated into Axial, but I was no longer viewed as a first time entrepreneur. I was someone who had experience. They didn't ask me to fill out a questionnaire of what I really knew. Just the fact that I had a company that had already raised money gave them confidence that I could not just replicate it, but hopefully exceed my contributions from the first company. I think that's a thing where people say, "Oh, this is your second company?" They just automatically give you the benefit of the doubt.

ZIERLER: Sarkis, on that note we'll pick up back on the science and how Axial got started next time.

[End of Recording]

ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Friday, September 9th, 2022. I'm delighted to be back once again with Professor Sarkis Mazmanian. Sarkis, once again it's great to be with you. Thanks for joining me.

MAZMANIAN: Hi, David. Thank you.

ZIERLER: Sarkis, today what I want to do is further develop where we left off last time. That of course is the origin story of Axial Biotherapeutics. I want to frame the question like this especially in light of where we were last time where you were emphasizing you were no longer a novice on the business side of things. You had learned a lot from the previous experience. What aspects of the origin of Axial Biotherapeutics were simply an opportunity for a redo where the science was really similar, but you could build a new company from scratch and pursue similar lines of scientific inquiry? Where was there really new science, new technology, new discovery that prompted the need for a new company regardless of the past experience?

MAZMANIAN: I think I'll tackle the second one first. It just feels a little bit more natural to me. Under a larger umbrella the technologies sort of are similar because they involve products of our research. Our research has its own focus and that's on gut bacteria. At that level we're still looking at technologies that emerged from the microbiome. What's different were indications—meaning the disease areas—and mechanisms of action. The first company, Symbiotix, was based on an immunomodulatory technology. The targets there were inflammatory disorders and autoimmunity, whereas the technology that was wrapped in to license to Axial were technologies more in the neuroscience space, more modulating complex behaviors or impacting the viability or function of cells in the brain. By nature they're different, but both in terms of what the disease areas are—for Axial it's now autism and Parkinson's disease—but also what the mechanisms are. For Symbiotix it was immunologic mechanisms; here it's neurologic and really at its core metabolic, but that's more non-obvious, I think.

ZIERLER: Why is that non-obvious—the metabolic connection?

MAZMANIAN: Because in the body everything is talking to everything. I think we've discussed that. In this particular case, while the target cells are neurons or cells in the brain that support neural function, the link between the microbiome and those neurons—microbiome being the gut, neurons being the brain—are small molecule metabolites that were produced from dietary substrates that mediate this crosstalk between the gut and the brain. Those molecules are not unusual to other molecules…are molecules that are produced by bacteria, but then are modified or transformed in the liver. The actual product is a result of co-metabolism between bacteria and us. Really it's microbial metabolism that feeds into human metabolism that creates a small molecule that affects brain cells. Metabolism is the link. It's non-obvious because most people don't think about what peripheral signals affect a neuron. It could be neurosignaling or it could be small molecules. In this case small molecules.

The indications are different and the technologies are different. In terms of what I learned in applying that to a second company…I think this is the power of, David, where your questions are allowing me to uncover things that happened that I didn't realize were happening at the time. I think I probably…I forget our conversations; that's just how I am. I probably discussed some of the shortcomings of Symbiotix and really the leadership issue.

ZIERLER: Sarkis, we talked about that in real time, but not from the postscript from reflecting back and thinking about how you could do better next time.

MAZMANIAN: That's right. I gave you the Symbiotix story, right? Unbeknownst to me there was a dynamic playing out—which I'll describe in a second—on Axial that really fed into this concept. With Axial when I first was thinking about the technology I knew it was significantly different both in the ways I've described. I viewed it as a different company and then I asked around to see if there was somebody who could help me from the business side build the company. Through connections with a Caltech trustee I was linked to somebody in the area here who had a lot of experience in entrepreneurism and really getting companies off the ground. His name is Michael Glembourtt. Michael and I started putting the framework for the company together—what would we be offering, what are we presenting to investors, and what the mission, the deliverables of the company would be.

I went out and started fundraising. This went pretty well, actually now that I think about it. We didn't talk to too many people because we got a lot of traction from investors who had worked with Caltech before. One was more of an angel group, a friends and family group, but the other was an institutional investor called Domain Associates. Our managing partner at Domain Associates is a person named Jim Blair who had funded a number of companies at Caltech. He's got this pretty good relationship with Caltech, but also an association with how things work at Caltech. I got a meeting through one of the Trustees. With Jim Blair, he just came to my office—without Michael—just me and him chatting. He made a comment; I forget the exact words. After an hour he said, "This is the kind of thing we want to invest in."

ZIERLER: Was Jim's experience in bio with Caltech or was he doing things across the board?

MAZMANIAN: He's in biochemistry. I think he's done some engineering companies as well. His fund is really a biotech fund. Even if it's engineering it's like medical devices and materials. I'll speed this part up a little bit. Domain was on board and then serendipitously I got a phone call from somebody who I had worked with through Symbiotix. His name is David Donabedian. David was the director of AbbVie Ventures. AbbVie is a pharmaceutical company. They have a venture arm; he was the head of that office. We had worked with him on a small research collaboration at Symbiotix. David had left AbbVie and went to a venture fund called the Longwood Fund. I should ask him, but I suspect he called a number of people that he had worked with over the years because his new role at Longwood was to find new opportunities. I was one of the people that he called. He said, "What do you have going on?" I told him about the company which at the time I had named Neurobiotix. It became Axial after they brought a positioning firm in to change the name.

ZIERLER: I assume that emphasizes the gut-brain axis.

MAZMANIAN: That's right. I think they paid $15,000 for the word axial.

ZIERLER: [laugh]

MAZMANIAN: I told him about the predecessor to Axial and I told him that Domain was on board. I linked the Longwood fund to Domain; they had never worked together. They seemed to get along even though there are sort of different philosophies. They basically gave me a term sheet where they said they were going to co-lead that Series A. It was a very sizable Series A; it was a $19 million initial investment in basically an idea that I had in some early patents, none of which were issued. Here's where the lesson part comes in.

What David in his role at Longwood that he had taken on was to not just fund companies, but to insert himself in the management team, almost serve as the CEO of the company to get the company off the ground. Previous to any of this, I was working with Michael. Fast forward to here, by the time we got to the term sheet Michael and David had met and Michael had met the folks at Domain. But David was very clear that their investment was contingent on replacing Michael with David which is not uncommon with venture funds. They want to put in their people.

ZIERLER: Sure.

MAZMANIAN: I hope that they would want to put in good people. I suspect most of the time they just want to put in their people. I think it was probably a combination of the two. I had to make a decision—there's the money component—but who is the leader of the company?

ZIERLER: Sarkis, how did you respond to that before even analyzing whether or not you do it? What did that tell you about the venture capitalists, what they wanted to accomplish? What did that signal to you?

MAZMANIAN: Without reading too much into what they're actually thinking—because that's a hard question to answer—I think I just thought about it pretty simply. Like I said, this is not unusual. As this is happening I'm talking to people like Ray Deshaies and others who are giving me advice. I learned that this was not unusual; it's not uncommon for it to happen. I didn't put too much into it. My focus was on really, Who was going to be the best CEO of the company based on what was turning out to be a negative experience with the CEO at Symbiotix? I spent probably a good month or so really trying to interview both for the role even though if I wanted Michael in the company, it would look very different than if I work with David in terms of the financing. I was less concerned with that, more concerned with who had the vision. The ability to execute that was going to be the best for the company.

I went with the deal that Domain and Longwood offered and we gave Michael a nice severance. I know he was pleased at the end of the day. He was disappointed he was not going to lead the company. I think he had bought into the technology and I think he wanted to work with me, but I know he was happy with the compensation he got. He built a lot of value in the company and he got it to the point of being funded. The lesson there is—I think this is so critical—just having the best possible management team, really the best CEO come in. They're all different both in terms of personality, in terms of how they want to build the company, what their vision is, how I work with them, and many other variables as well. I believed at the time, and in retrospect I now know David was a good choice. I can't tell you who was a better choice, but I know he was a great choice for the company. The company did well in its early stages and then continues to flourish to this day. Again, the take-home message is a good management team to the best of the ability that any entrepreneur could ever identify and/or recruit because sometimes the best people aren't going to come. Most of them aren't going to come to early stage companies. Here the stars just aligned well.

A New Venture with the Benefit of Experience

ZIERLER: Sarkis, in the launch phase, thinking back to your first experiences—going from the first to the second company—did you ever consider even for a second leaving Caltech so you could put your full resources into it?

MAZMANIAN: For a second, but it was actually not until my third company that I thought about that. Not at this point, not for Symbiotix, not for Axial. Maybe there was a second; I can't tell you for sure that there wasn't. Whatever it was it didn't imprint too much into my mind. I guess it was just one of many possibilities. For Axial it was not on the table.

ZIERLER: Is part of that the confidence and understanding that you had in getting a quality management team in place so that doing something like leaving Caltech simply wouldn't be necessary for you?

MAZMANIAN: It was more along the lines of I love what I do here and I didn't want to do anything else. After the company was founded the idea was floated to me to essentially come on board as a CSO. It was a very casual ask by Axial, but I think even they knew that for an early stage company why would I leave a place like Caltech with job security and I'm doing what I'm doing for that particular opportunity. I guess from their perspective they thought it was what would be best for the company and they should ask. Obviously I have an emotional attachment to the technology and to the success of the company, but that was enough. To whatever degree that proposition was made to me, I think both sides realized that I wasn't going to do it.

I'll fast forward really quickly to Nuanced. It's the most recent company where I did really consider and I even spoke with the investors prior to financing the company the possibility of taking a sabbatical—it wouldn't be permanent—and either being the startup CEO/startup CSO of the company. When I really sat down and thought about it the consequences of doing so would mean that our research would suffer. That to me was just a price I was not willing to pay.

ZIERLER: You're thinking about your grad students, your postdocs, keeping the lab going.

MAZMANIAN: That's right. If I take six months or a year away from the laboratory it's just really going to impact what we're trying to accomplish here. Once I thought about it on that level…I think it's hard to see the lab maintain its level of success without me being here. I'm not trying to be immodest. I just think I have a role; the students and postdocs—they have their role. I certainly didn't pursue that. Parenthetically I've been here 16 years as of September 1st actually. I think I might have told you, in 2006. I still haven't taken a sabbatical all this time because I worry how that's going to affect our work. [laugh] I realize that that's just what matters to me. I think our academic output, our training and working with students and postdocs—this is what drives me and motivates me. Any scenario that takes me away from that even temporarily is one that's very hard for me to consider.

ZIERLER: Sarkis, let's go back now to the science. The question is about the role of scientific publications in the launch phase of a company where you're just presenting fundamental research to your scientific community, not necessarily to the investor crowd. What is the role of publications when you're thinking about the launch of a new company?

MAZMANIAN: I think whether it's at the launch stage or even once a company is founded, there are a number of things that are considered stamps of validation. Scientific publications are certainly one of them. The vast majority of the investors that I deal with have a Ph.D. Those that have an M.B.A. or other backgrounds are usually also conversant in science to some degree. I think that just from that perspective they revere any discoveries that were peer-reviewed and really validated by the scientific community. Most investors would go out and try to get more money. The name of the game for companies is to raise money and then turn around and raise more money. They have to always convince more and more people. Every single company that I've been involved with, and the vast majority of companies where I've seen the presentation, their slide deck, they highlight their scientific publications. It's really a metric that everyone I think is looking for both at the launch, but also as the company continues. Just one real world example. Every paper that we publish from our lab where the technology overlaps or supports Axial's thesis, Axial will have a news release on those papers. They value the scientific publications.

ZIERLER: Either with specific papers or more generally the ideas contained therein, what was some of the research that was most influential, gave you the most confidence in the launch for Axial?

MAZMANIAN: There are two studies. I can really pinpoint it down to two papers. One was a 2013 paper that we had continued to develop where we showed that a particular probiotic was able to ameliorate the behavioral symptoms in a mouse model with autism…is the work that we did with Paul Patterson. The second was work that we were doing in Parkinson's disease, because as we got really encouraged by the data we were seeing in autism, we expanded to Parkinson's largely because of the gastrointestinal symptoms in individuals with Parkinson's. It's pretty striking. Eighty percent of Parkinson's patients have constipation prior to their Parkinson's diagnosis. It's just part and parcel of the disease if you will.

I thought this was curious and so we had started this research program on Parkinson's and had made a lot of advances. When I was fundraising—this was end of 2015, early 2016—we had the data on Parkinson's, but essentially we knew that not only that the microbiome was involved, but we had some candidate organisms that we felt were pathogenic. Meaning that they were the ones that may be promoting Parkinson's like symptoms in mice which was eventually published in a nice paper at the end of 2016, that data package. Finally the paper that came out which was very close to the time we launched the company really spring boarded both the financing and the publicity for Axial around the launch. These are papers that the company to this day highlight as part of a scientific basis for the central focus of the company.

ZIERLER: Sarkis, from the first venture looking to Axial, did the regulatory framework change either because this is simply new science or because the policy changed? Or simply your experience grew? You knew how to work in the system better?

MAZMANIAN: In regulatory you mean like with regulators?

ZIERLER: Exactly.

MAZMANIAN: With the FDA. My sense is the FDA has been and continues to be very bullish on technologies coming from the gut microbiome. They've stated this to me in really no uncertain terms that their position is…because the FDA is concerned with safety and tolerability of drugs, not whether they work or not, but are they going to have any negative effects? They're convinced that if the drug—which is in most cases a bacteria or bacterial molecule—comes from the human gut then evolution, civilization has done the toxicology studies. Taking a bacteria from one human and putting it into another human is a very low risk proposition for them. They have both stated and ultimately acted on this position that they are going to be very supportive of technologies that are microbes or their molecules coming from the gut. The evidence for that is a number of workshops.

There was one year I went out to the FDA three times—they invited me out. Essentially it was learning lessons for them just to sort of educate them about what's going on in the field. The real sort of proof being in the pudding is that Axial has gone to them now with a couple of IND filings and they've been…I'm not the world's expert on getting IND approvals from the FDA, but from what I've heard from our regulatory team it's been one of the easiest experiences for them. You're not trying to convince the FDA that this small molecule made in a laboratory is not going to be toxic when you put it in people. What you're saying to them is I'm taking this bacteria from a gut of one person and putting it into the gut of a different person. That process has been very, very smooth. Both INDs that Axial has applied for were not just approved, but from what I hear approved in a very streamlined and easy process. I think that's their position and I suspect that that's still maintained.

Just in the field we've heard only stories where microbiome based technologies were safe. A few examples—very isolated, limited examples—where there were some adverse events, but again hundreds and hundreds of studies now in people with very minimal hint of any adverse events. So, I think the FDA is pretty well sold on microbiome based technology.

ZIERLER: Sarkis, are there any genetic issues to consider, when you're extracting material from one person and putting it into another? Does it affect anything at the DNA level?

MAZMANIAN: DNA of bacteria or DNA of humans?

ZIERLER: Either one.

MAZMANIAN: It's unknown, I would say. Probably there's a likelihood that it could affect both, more likely that it would affect the DNA of bacteria. This is all speculative just to answer the question. I could see a scenario where mutation rates or other genetic changes, horizontal gene transfer, those rates could be different in an organism depending on whether it's in one person versus another. To my knowledge that's never been looked at. It's an interesting question. Even more tenuous would be would any microbiome based intervention or change affect human genetics or DNA? I think that's probably unlikely unless it's some increased…this is maybe the most obvious of the very tenuous outcomes, but is it going to cause mutation rates to increase in intestinal epithelial cells that may lead to colon cancer or something? But, yes, I think the answer is pretty firmly…no one has actually looked into those questions.

ZIERLER: It's clear though between the FDA's position and the safety record of millions of years of evolution the unknown itself is not a concern.

MAZMANIAN: Yes. If you take fecal transplants as the test case for whether or not putting microbes…and whether it's directly to the molecules or if you put the microbes and they're going to produce some molecules from one person to another is harmful. There's like 12,000 clinical reports of fecal transplants. Again, no real issues of safety. Of course, the FDA sees these studies as being quite safe. They sponsor many INDs. What we're talking about here in terms of both what Symbiotix was trying to accomplish and what Axial is accomplishing is essentially to put bacteria in the molecules or drugs that affect those molecules into humans. The experiments that have been more likely done is I don't know which bacteria is therapeutic. I think that there's therapeutic bacteria in one person; I'm just going to move all the bacteria because that's just fecal transplants. If you move everything from one person to another, that's been done at least 12,000 times. That's been clinically documented. There's a lot going on that's not been clinically documented. It's very, very safe. I think moving forward I don't think the FDA is going to have much of an issue.

ZIERLER: Sarkis, just a mechanics question in creating the medicines that Axial wants to produce and market. From extracting it from the microbiome what does that look like? How do you store it? How do you manipulate it? How do you mix it into a medium that can be ingested by patients?

MAZMANIAN: It depends on what it is. Is it a live organism? Is it a component of the bacteria itself? Is it a small molecule or any other molecule that's released by the bacteria? I think that's going to be on a case by case basis. I wouldn't say it's been fully solved, but in most cases I think the extraction and storage components, you can overcome any of those issues. It's going to depend on the organism, the nature of the molecule, but those are solvable problems. The last part of your question was what, David?

ZIERLER: Storing it, manipulating it, mixing it into a medium that can be ingested by patients.

MAZMANIAN: The ingestion part…the three approaches are if it's a fecal transplant then it's done by enema or nasal pharyngeal tube. That's one, so you can think about that—pros and cons. Obviously pills. There are people who do fecal transplants and take dozens of pills. Or you can put one particular…like a probiotic they would get at the store. You just put it into a pill. That's the second. The third is something that Axial have been using in their clinical trials is essentially where the drug comes in a sashay and you essentially mix it with your favorite food because it's not amenable to putting into a pill. Or in Axial's case the patient population is unable to take pills. Autistic kids generally have a hard time swallowing pills. You can give it with their favorite food.

I have not yet seen any microbiome based therapeutic that was injected, where the mechanism is to inject. I think that would…I'm not saying that it doesn't exist. I haven't seen it and in fact if it did I would say that they are sort of not really harnessing the potential of the microbiome. The whole point of the microbiome is that it works in the gut. [laugh] Why even think about doing things like injections? If you go out into the marketplace people like to take pills. It's easier to take pills. People don't like injecting themselves, so you shouldn't go down that road unless you absolutely have to. Again, I haven't seen those technologies.

ZIERLER: To go back to the first venture where there wasn't really a brick and mortar component to the company. Was that important for Axial to have a building, to have your own infrastructure if you will?

MAZMANIAN: Yes. Axial made its own decisions. To whatever degree my experience is that Symbiotix shaped those decisions were probably minimal. When the CEO and a good management team were placed decisions were made because of what's best for the company. There was never a discussion about keeping Axial virtual, because what we had to accomplish just required scientists. It required laboratories because there was work to do that couldn't be done in my laboratory. It required animal models that didn't exist at commercial research organizations. I think the brick and mortar model was probably the only viable one. Again, not much discussion on a virtual company there.

The one thing about that was just because brick and mortar…even though a lot of companies were going virtual even well before the pandemic—because the commercial research organization field is just exploding…this is a fun fact; you're probably not interested in this. More money is now spent and has been for several years now…Pharma spends more money outsourcing research than they do in their internal R&D programs. These commercial research organizations, they take business not just from academics or biotech, but from Big Pharma as well. It's how business is being done. I still think the vast majority of biotech companies are brick and mortar. There are companies, scientists, and people. I think that probably influenced it, but if you just look at what needed to be done, it was clear that we needed an actual space and scientists.

ZIERLER: Just to give a sense of the projected size of what Axial would become were you looking at, was the CEO, looking at buildings? Were you looking to rent a floor in a biotech kind of building? What were the options? What did you need?

MAZMANIAN: I can just tell you what happened. It was kind of a slow build because we were looking for space. We spent a lot of time looking for space here in Los Angeles. Domain is here in Southern California. Investors like to be close to their money. Longwood is in Boston; the CEO is based out of Boston. We spent a good six months while we were still putting a lot of pieces in place looking for a physical site here in Los Angeles, and it just didn't work out. Really the rate limiting step was the lack of infrastructure to house and study mice. Where are we going to do the animal work? As soon as we started looking in Boston, it was just a whole different ballgame. There's just such a huge community in biotech, and these support CROs which are just ready to ramp up any project when you do mouse work. Under contract, let's say a month from now, whereas if you were going to build your own vivarium and your animal work it would take a year and a half, two years to do the same experiment. Once we started putting those calculations together I think it became pretty obvious that having a home in Boston made more sense. And just to be candid, there's simply more talent in Boston than there is in Los Angeles in the biotech space. So, we founded the company there. Initially it was an office and then we rented lab space. We leased a few bays in our larger lab where there were many other companies as well, but within six months we had our own site. The company has had its own site ever since. It's moved a couple of times. About a year ago we moved to a new site that we built out entirely ourselves.

Building Los Angeles into a Biotech Research Hub

ZIERLER: The larger story here—I know this is something David Baltimore is working on—is that Los Angeles still really is not a biotech hub even sandwiched in between San Francisco and San Diego. If Los Angeles did have more of a biotech component would that have made staying close easier do you think?

MAZMANIAN: Absolutely. In fact that's what happened with Nuanced. It is located in Los Angeles because a couple of things happened. In the interim between 2016 and 2020 there was an expansion of capabilities and facilities here in Los Angeles. I also participated in some of those initiatives to build out more biotech infrastructure and to keep talent here in Los Angeles as opposed to having them move to other biotech hubs. I'm really excited about the future of this industry in Los Angeles. If I could start every company here in L.A. that would be my preference.

ZIERLER: That's better for you, obviously.

MAZMANIAN: It's better for me, but it also depends on the investors. Again, investors want to be close to their money. The other aspect is in some ways the space component is the easier one to build up—buildings and vivarium. Really that network of individuals because there's so much turnover in biotech at the executive level and just the team level…if you're in Boston, people are just always hopping from company to company. There's no shortage of talent. I still think it's going to take some time to build the human resource and talent base here in Los Angeles where you just have…you want to fill a role? There's a number of people to choose from. We're nowhere near that, but I would love for that to happen. I'm hopeful. I think there's a lot of people who are quite committed to really expanding the biotech, the biopharmaceutical industry in Los Angeles.

ZIERLER: Between UCLA, Caltech, and USC, just from the academic talent level are we at a critical mass point now where Los Angeles has enough people where this just simply has become more feasible?

MAZMANIAN: I think the people are here. In terms of the way you described from the universities this is the talent pool that's being trained. I think L.A. hopefully is recruiting talent. That people who, let's say, are coming from Boston, San Diego, San Francisco, what have you, they're moving to Los Angeles. Really you want to be able to have a pool of executives to choose from as well. Those people are probably going to have to relocate; a lot of them are going to have to relocate. Feeding the funnel in the long run…I think there's talent everywhere and in a large city like L.A., there's always going to be enough people to staff these companies. I just think you get to an inflection point where there's a critical mass and I don't think we're there yet. The first steps are to build the physical infrastructure, which is happening. There is incubator space going up all over the L.A. Basin and all the way out to Thousand Oaks. There's a lot of activity going on around Amgen. More and more investors are opening offices in Los Angeles. I think there is clearly momentum moving in that direction.

ZIERLER: Sarkis, you mentioned physical space, lab space, animal research. At what point do you start thinking about a clinical infrastructure, having access to patients and hospitals? Is that a consideration right from the beginning or that happens in due course?

MAZMANIAN: It happens when the need arises. That's where you don't build that infrastructure. You plug into a hospital. Axial is currently recruiting for a 195-person trial now in 21 sites between New Zealand, Australia—which have four sites—and the 17 sites here in the U.S. The clinical component comes in when you partner with a hospital, more specifically a clinic within the hospital, more specifically a clinical principle investigator within a medical center, where they have both the infrastructure to run clinical trials, but also a patient population to draw from to enroll in your clinical trial. Once you get to that point, it's different for different diseases where in some cases there are larger populations, others are small populations. What I've also learned that's kind of interesting is there is competition for patients. There's multiple clinical trials for the same indication. You're competing for patients because obviously they can't be in two trials at the same time. All that is plugging into existing clinical infrastructures.

ZIERLER: Sarkis, just to orient me in the chronology—when is the launch of Axial? When does it become a real company that's on its own?

MAZMANIAN: November 14th, 2016.

ZIERLER: Wow.

MAZMANIAN: For these companies there is a founding date.

ZIERLER: Right.

MAZMANIAN: A lot of stuff happens, lots of conversations, lots of paperwork, but there's an actual calendar date where the company was founded.

ZIERLER: The deeper question there is does that date, is that a sufficient stand-in for you when it feels real after pouring your heart into it for all the years leading up to that moment?

MAZMANIAN: It's like marriage.

ZIERLER: [laugh]

MAZMANIAN: Nothing changes. [laugh]

ZIERLER: That's great.

MAZMANIAN: You've already committed to this many, many months ago. You know it could fall through obviously, but by the time you get to that founding date the work has been done and you've already mentally and emotionally accepted that this is going to happen.

ZIERLER: For you, what were the pros and cons of obviously being here in Pasadena with the company in Boston? What worked? What was difficult as a result?

MAZMANIAN: I don't know if there is any sort of bright side or upside to being physically distant from the company, but it wasn't a hinderance to the company either. Because the way Axial was built, it was going to outgrow me at some point. In fact in some ways I thought it would outgrow me by now and it hasn't, but there are very positive reasons for that. I just have a really good relationship with that company. It's not unusual for an academic founder to be phased out after two or three years in the company. We can discuss that too; it's an interesting concept. I don't think it hurt the company that I was at a distance. Between teleconferences—I was doing Zoom. Did I tell you this? Yes, well before the pandemic.

ZIERLER: Right. You were well prepared for the pandemic.

MAZMANIAN: The CEO was very good initially just about coming out here to Los Angeles. All of our first board meetings were here in L.A. Once the company started really…had a footprint and grew in Boston and all the subsequent board meetings have been there. Before the pandemic I was out there at least four times a year, usually more because I was going out for board meetings, but I'd go out for other events as well. We'd have a science day, or if there was a critical meeting with a pharma company or a venture fund I'd go out for those as well. Oftentimes I'm in Boston for other reasons, conferences or to visit some of the institutions there, so I'd spend a day at Axial. I don't think me being at a distance has hurt the company at all.

ZIERLER: Between your conversations with the board and the CEO, what was the sense of timelines, of benchmarks for success, for getting a trial done, creating a medicine, working through the process? What are the months, years, and even decades that you're thinking of circa November 2016?

MAZMANIAN: I'm not sure I can answer that particular question, David. I suspect that there were timelines, but early on in the company I think those are a lot of times placeholders because it's just hard to know whether there are going to be obstacles or delays or even accelerators.

ZIERLER: Sure.

MAZMANIAN: I think those timelines early on, which I'm not even calling them at all…I think it's just due diligence. You just have to put those slides together. This is the timeline, but I think everyone takes those with a grain of salt. So, I can't answer those questions. I think the bottom line is sort of the more generic response is you want to just get everything done as soon as possible.

ZIERLER: Yes. So, almost eight years out looking back what benchmarks have been reached even if there wasn't a clear sense of what they were or what it would take to get there?

MAZMANIAN: Six years out, right? 2016.

ZIERLER: Oh, 2016. Right!

MAZMANIAN: Looking back I think the company has moved very, very fast. The company made some very good decisions in accomplishing the goal of let's just take this technology in humans as soon as possible just to get a signum of whether or not it's working. Here's more of what I learned with this experience that I never got a chance to learn at Symbiotix because we didn't get that far. This idea that a clinical trial is an experiment. A clinical trial which you expect especially early on in the life of a company is to inform your next clinical trial. It's not to lead to a drug; that's a lofty goal. Axial under the directorship of the board—I think this was really something the CEO embraced as well—said is what we're talking about here is unproven technology. It's completely blue sky. It works in mice; that's great. There's a lot of things that have worked in mice that never worked in people. [laugh] We can spend years and millions of dollars doing experiments in more animals or in vitro just to give us some increased confidence that it's going to work or not, but ultimately you've still got to test it in people.

The plan was to get this into humans as fast as possible. Just to get a safety signal, a tolerability signal, maybe an efficacy signal as opposed to spending the capital and the time to get us to a point where the research community would never reach conviction. You might reach with confidence, but not conviction. The company just said, "What's the critical path to getting this technology into people?" Less than three years after the launch of the company, we were dosing humans in a patient population with the drug that the company hopes to develop, hopes to even commercialize someday. That's important. The company made some decisions where the drug didn't go in—because this is not uncommon…where the drug initially goes into healthy, normal volunteers. Because at the initial stages of any clinical trial, this is where to look at safety and adverse events, and there you start with healthy adults.

In the company's discussion with regulators, I think they reached an agreement that this drug could be perfectly tolerable in a 35 year-old healthy adult, but still may cause adverse events in an adolescent with autism. So, you're not going to learn anything from the healthy human, normal volunteer study. Go directly into a patient population. Which is like a win-win, right? As long as there's no side effects, it's a win for the patients. It's a win for the regulators because it advances the program in their minds, and then obviously a win for the company because less capital is spent on another trial. You go directly into your patient population. Within three years we were dosing adolescents with autism with an experimental drug. I think that's, from the best knowledge that I have, that tells me that's lightning fast.

ZIERLER: Yes.

MAZMANIAN: I give the company a lot of credit for that.

ZIERLER: Sarkis, what specifically…it is lightning speed in biotech. These things can take much longer than that. What do you credit specifically? A solid CEO, but what else makes this happen in three years record time?

MAZMANIAN: A number of things. The support of the board. Everyone buying into this vision. And then company scientists who are going to figure out the how. If the goal is to get this into people as fast as possible, well how do you do that? How do you identify the therapeutic that you are both able to convince others will be safe and then manufacture it and dose it in ways where you're going to be able to detect any of those events? What the company did which I thought was very, very smart was, it said, "We want to accomplish a particular chemistry with the drug." You can do this with a bacterium, you can do this with an enzyme, you can do this with a small molecule. The first decision that the company made was to move away from the bacteria—and I can talk about the reason which I think did, was lowering levels of this microbial molecule. You can do this with bacteria. Well that's a long road because no one's ever commercialized the bacteria. You can do this with an enzyme or let's say even a small molecule, but if you do this with a novel compound then you have to go through really rigorous safety studies that take many, many years.

Or—and this is what they ultimately did which makes sense—you can repurpose an existing drug. If you take an existing drug that's already been approved by regulators for, let's say, one disease or one patient population, and either you modify it or reformulate it, it's a much faster path into the clinic because the FDA is already familiar with the safety profile of that drug. It's got to do something either in a different population or the chemistry has to be slightly different than the existing drug, so it's going to accomplish a goal, an objective that you've sent out to accomplish.

This is what the company did. They actually found a drug that's made by a Japanese manufacturer. It's only marketed in Japan. It's marketed for chronic kidney disease, but they realized that if they take that drug and reformulate it a certain way that it's going to have the activity that we want that we predict would be protective in autism. With that path, as you can imagine, to even identify the drug a number of decisions were made that this is going to get us into the clinic as fast as possible. It may not be the ultimate drug. It may not be the final drug that you make, but if you see a positive signal in the clinic then you know you're on the right track. The concept that a clinical trial is an experiment to just inform your next experiment—this is a trial or an experiment—but it tells you that your idea which has been validated in mice may actually work in humans. That's of huge value especially when you're going into a space where there are no drugs approved for autism or for the core symptoms of autism.

Lots of people are skeptical that you could ever develop a drug. It's not like the 14th cholesterol drug. If you're developing the 14th cholesterol drug a lot of people will believe you can do it. If you're developing the first autism drug there are a lot of people who will tell you, "It's never been done before, so you cannot do this." Just getting that initial proof of concept has huge value to the company. That's what the company did. They repurposed a drug. We had a very limited market. They were very smart in how they negotiated the business. That company—again a very small pharmaceutical company in Japan—really didn't have the ambition I guess to take their drug and market it for other diseases worldwide. So, Axial owns the rights to market this drug anywhere outside of Japan. If I were that company, I would never have done that deal that they did. [laugh]

ZIERLER: [laugh]

MAZMANIAN: But they did. Now Axial's reformulating it and coding it to be released in the right part of the gut and there are all sorts of new patents we're going to write on its formulation and its release because it changes its efficacy and safety profiles. Long story short, that's the route that we took on a series of very wise decisions with the goal of saying let's test this concept in people as fast as possible because that's going to lead to more buy in that we're on the right track.

ZIERLER: Sarkis, last question for today. In comparing your early assessment, thinking about Symbiotix where you indicated there was a little voice in your head even though you were new to this world. There was something telling you you had concerns even from the beginning. Did those melt away with Axial? Had you known enough? Was the company in a situation where looking back you saw that it was on a path to success even if you couldn't obviously project into the future?

MAZMANIAN: Yes. I'm never sure I'm confident about anything.

ZIERLER: [laugh]

MAZMANIAN: I guess I was more comfortable with the path that Axial was taking. I think a lot of it was the learnings from Symbiotix. Learning isn't linear. It's not as if you learn one incremental amount of information each day. [laugh]

ZIERLER: Sure.

MAZMANIAN: I think I've learned a lot, to be honest with you. The rate of learning has slowed down, but I'm always learning. I'm also learning very different things because activities of Axial are very different. Symbiotix never got this far. I felt like both the experiences and my intuition and my relationship with David and the rest of the company gave me just a lot more comfort that Axial was on the right track. Decisions were being made for the right reasons. Everyone was pulling in the same direction. The people were talented, working really hard. Nothing is guaranteed, but I kind of felt like all the right pieces were in place. I have concerns about Axial because I have concerns about everything because nothing is ever finished. It always feels like they're not insurmountable.

ZIERLER: Sarkis, another terrific conversation. Next time let's pick up on some of the science and mysteries surrounding autism and where you saw Axial slotting in.

[End of Recording]

ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It is Tuesday, November 29th, 2022. It is great to be back with Professor Sarkis Mazmanian. Sarkis, at long last it's great to be with you again. Thank you so much.

MAZMANIAN: Hi, David. How are you?

ZIERLER: Great! Sarkis, today we are going to pick up really right where we left off last time. It's about the connection with Axial Biotherapeutics and autism. Before we get to the specifics of that just to zoom out a little bit…what is the larger framework by which you identified autism as one of the CNS issues for which a startup would be particularly well suited to tackle?

MAZMANIAN: I think for that it really was the strength of the data, and I guess the novelty and excitement, but with that comes risk. Really the strength of the data where we were and remain quite convinced, at least in animal models, the microbiome has a very profound effect on behaviors—those that coincide with human autism as best as it can be modeled in mice. Also other features that are comorbid or associated with autism like anxiety and depression, where again not only our work, but many other laboratories have shown these very effects by manipulating the gut microbiome. Once you have a certain data package and if it's compelling…and again luck has a lot to do with it and being the shiny new object has a lot to do with it as well. This was 2015-ish. We'd already published several papers on the connection between microbiome and autism at the time. I felt like the data was strong enough and there is a medical need obviously in autism where there are no drugs to address the core features. SSRIs are given not just to individuals with autism, but many people with generalized anxiety as well. So you can treat symptoms, but not the core symptoms. The other caveat, and I'm happy to get into this if you want to, is just this whole discourse on whether or not autism should be classified as a disorder or a disease rather than just on the neurodiversity spectrum. I have my thoughts on that as well, but it's separate from the point about Axial.

In terms of the mechanics of that—I don't know how much of this we went into—the fundraising was what I would say pretty traditional in the sense that I put together a pitch deck, I leaned on some contacts that I had made through Symbiotix that helped me really think about the business. A couple of people that I brought from previous relationships, but also through introductions who could help run the business and make introductions to ambassadors and things like this. And, how do you really pitch a story to investors? Really it was two things.

There was one investor named Jim Blair who has a lot of associations with Caltech. He's started a number of Caltech companies. He's at Domain Associates, which is the fund that he runs. He's the managing partner and he's been there I think many decades. He got very excited about this and that was through a Caltech introduction by Rich Wolf, who was a trustee here. When you have someone like that excited I think other people tend to get quite excited as well.

ZIERLER: Did he have a family connection? Did he have some personal connection to suffering with autism that got him excited?

MAZMANIAN: Not to my knowledge. I remember the actual dialogue. Again, I had brought aboard someone named Michael Glembourtt who was a startup CEO. The understanding was that he would help get the company to financing and then would step away from the company; he was semi-retired already. Rich walks in with Jim—it was a scheduled meeting—it was just the three of us. None of the other people were there. Jim just got really excited about the story. It was a pretty basic one hour meeting. I don't know exactly what it was, but something resonated with him to the point where he was excited about investing in the company. Within a few days it was clear that there was interest both from him and the fund—I went down and I pitched to the larger fund.

He got excited, but then the other aspect was, just as we were putting the company together our lab published another major paper on the connection between the microbiome and Parkinson's disease. It was one of the first papers—maybe even the first paper—on showing the mouse models of Parkinson's are dependent on the type of bacteria in their gut to resolve or exacerbate motor symptoms. I mentioned to him this data package. It was not part of the presentation. I said I was open to having that be part of the company as well. I think that that helps de-risk a company like this because you have multiple assets, multiple different disease indications—or at least more than one. It helps offset that risk of really having wins for the company or having a trajectory for the company that isn't challenging. The reason why that was an enabling event…the timing just worked, and folding that technology into Axial just made sense because they're both sort of in the gut-brain space. The fact that there are no drugs for autism…though to the naïve person like me at the time I thought to myself, "What a great opportunity." To drug developers, they think, "How risky!" [laugh]

ZIERLER: [laugh]

MAZMANIAN: We've never done anything to move the needle, certainly not in a disease modifying way, but not even in a sort of interventional or therapeutic way that's core to the features. That just gives it a lot of angst whereas with Parkinson's there are treatments, so there is more confidence that you can develop novel drugs to actually improve people's lives. That aspect I think really enabled the company to come together. It was pretty painless as far as fundraising goes, where within a couple of months of that first meeting with Jim we had already rounded out a syndicate of investors. We had a target goal for the Series A. We had a straw man organizational team, identifying people who would come in as not just CEO, but head of research and legal and manufacturing and regulatory and things like this which are required for rapidly advancing any technology—not all investors bring know-how to the table. Some just bring cash; others bring teams of people. There's a variance and everywhere in between, but the investors that we had were very good about taking an active and human role in company building.

ZIERLER: To go back to what the data was telling you. Is there a debate in the field? In other words, are there people who argue that the source of autism comes from elsewhere that's not from the gut? Or is the idea that the gut is the source of autism, is that really the prevailing view at this point?

MAZMANIAN: It's not and it's not my prevailing view at least anymore. My thinking has evolved over the years as well where there was a time—and I think I both got caught up and contributed to the hype of the microbiome a decade ago—where it was supposed to be the answer to all of biology's unsolved mysteries and it certainly is not. I wrote an article a few months ago that was published on this point about the fact that this whole notion that changes in the microbiome may be causal to disease or even conditions or states, is probably not accurate in the sense that the effects of the microbiome are not as prominent as, let's say, genetic mutations. One could think about interventions like probiotics or fecal transplants and all the things that we currently can do with microbes and the effect sizes are nowhere near as potent as let's say a drug or many drugs. The question is where does the microbiome fit in either with causality, which is a high bar to reach, or contribution, or a consequence of the disease state. A lot of things can change the way the gut moves or the immune system or diet that then can impact configuration of the microbiome.

Long story short is my thought—and it's not specific to autism—but I think many disorders where the microbiome has been indicated is that the microbiome is a modulator of another risk or stress. Again, a genetic predisposition or some environmental stress or even some chemical stress that can alter physiology if your microbiome is "healthy." You're more resilient to being able to deal with that stress and maybe to the point where you may have—and there are lots of examples about this, actually—a genetic predisposition, but your microbiome is healthy enough that it masks the consequences of that predisposition to the point where you don't even manifest symptoms because your microbiome is robust and it keeps feeding your body with all the cues that it needs to function properly. But if your microbiome is altered—again it's a lock and key for all of us. There's no such thing as the healthy microbiome; we each have an optimal healthy microbiome for each of us which may be different within the population. If you have any deterioration or damage to the microbiome itself and that erodes the ability of the microbiome to provide resilience to the organism and now that initial trigger if you will…will manifest and may put you across a threshold where you are diagnosed with a disease.

This framework that I just described and the way we are thinking about this is now translating into some of our experiments where we are not looking specifically at the role of the microbiome in causing disease, but rather looking at gene environment interactions where the mouse has a genetic predisposition. Then, by tinkering with its microbiome we can either mask that predisposition through resilient effects of the microbiome, or if the microbiome is not providing all the beneficial signals or cues that it possibly can, now that animal starts developing symptoms. We're generating data to support this hypothesis.

Wide Open Questions in Neurological Disorders

ZIERLER: To go back to the idea that a decade ago there was hype around the microbiome being the be all and end all as it related to autism, but even more generally, what were some of the countervailing views at that point? People who were pushing up against the hype saying, "No, no, no. Autism is not connected. It's not all about the microbiome." If not the gut, where? Where else would autism come from?

MAZMANIAN: Genetic predisposition. The fact of the matter is that here in 2022 we don't know what causes the vast majority of autism. We can probably explain some small single digit percentage of autism based on highly penetrant mutations, meaning that it was in the germ line of a pair that got transmitted to the child. The child has that mutation, it manifests disorders. You put those same mutations into mice and the mice profoundly, very robustly, show symptoms. Those are causal mutations. That explains one, two, three percent of autism. The vast majority of other autism we just simply don't know. What people can find are when you do very, very deep sequencing of DNA, you can find that there are certain types of mutations, not in individual genes, but classes of mutations. These are called the de novo mutations which are more enriched in the autism population than the general population. People very much focus on genetics. There are a lot of people who are now pushing back on all the genetic research that was done to suggest that not only is it maybe not on the right path, but even if you do find causal implications of the genome what can you really do about it? Can you really go in and fix a mutation? A lot of people are trying to develop those technologies, and I don't think they're anywhere near ready for at least large-scale deployment.

At the time it's not as if people could point to—and they still can't—no, it's not the microbiome; it's this. The pushback which I think was founded at the time was based on an absence of data and more the notion that the microbiome cannot be this important. Some alien set of creatures in our body cannot be this important for our bodies, the way our bodies function. I guess you can debate. It was really along those lines because there is no solid evidence to say that it isn't.

I think autism—maybe I'll touch on this now—is a particularly sensitive area to study. You're talking about what is really an obscure disorder not just in the etiology that we talked about, but in its manifestations. There are so many different forms and features of autism where each individual is really different than any other individual with autism in certain ways, though there are certainly common features as well. Also you're talking about children and you're talking about parents who are rightfully and understandably very emotional. It's a hot button issue and there are people who take very strong positions on what it can and cannot be. You can go online; you will find battle royales going on online among people—again, there are vaccines and toxins and all these things that are being debated. Just to the microbiome there are people who will tell you the microbiome has nothing to do with autism and there are people who argue just as strongly that the microbiome is the answer to understanding autism. In fact, neither one knows even though they take these very strong positions. [laugh]

There's another issue that I'll layer on and hopefully this isn't too much information. There is this internal debate —regardless of the microbiome or any genetics or anything else—within the autism community of, Should it even be classified as a disorder? There are a lot of people who are very vocal online and on social media and they now attend conferences who say, "I'm autistic. There is nothing to study about me. I'm not different than you. You may have more hair. You're blond and brown haired, but that's just the difference between me and you." Those are people…I can see their position because they are able to function perfectly fine in society. Why would you put them in a disease group? What that doesn't capture—this is still being worked out by the community—are individuals who are—and it's now a buzzword—profoundly autistic, who cannot take care of themselves, who clearly have aggressive, self-injurious behaviors, who will need 24-hour care their whole lives. I think because it's such a heterogenous disease and you have populations of people who are highly functional and those that are not it just creates a tension when you're studying that disorder. Is there a patient population? Who is the patient population? How are they similar or different biologically, and can you address that?

I'll go back to your original question. I think the investors took a huge risk because autism is not a low hanging fruit in terms of a space to develop drugs, let alone with this novel concept that comes out of nowhere and is not based on 20-30 years of evidence to get to the point of, here is the intervention. I don't think much has changed in the six years that Axial has been alive. Even despite the success of the clinical trials, I think there is still a huge knowledge gap in to what degree does the microbiome contribute if at all to autism.

I'll go back and just maybe contextualize what I was saying earlier. If we can intervene in a person's microbiome…the hypothesis is the microbiome is not causal, but it's contributing to the behaviors in some way. If we can then restore healthier microbiome function, maybe it will overcome whatever that underlying condition or etiology is. Let's say for argument's sake it's genetics. That's where I think the company is now focused. Thinking of the microbiome as a modifier of behaviors, not as a cause of anything. If you think about it that way multiple things can contribute to disease. We're just looking at the one that is most actionable, which is the microbiome as opposed to let's say the genetics of an individual.

ZIERLER: Sarkis, I wonder if you've ever taken an anthropological approach given that the gut is so affected by diet. Are there cultures around the world who have a particular diet that makes their microbiome healthier and we see lower rates of autism? Have you ever tried to suss out those connections?

MAZMANIAN: I've thought about this a lot. Others have thought about this as well. Autism would not be the testing ground for that because to find those societies that have different microbiomes you really have to go to extremes. When you go to developing societies or even communities it's difficult to reliably diagnose autism. There's no blood test; there's not a brain scan for autism. It's really a behavioral questionnaire or a set of questionnaires, and I can tell you this through Axial's clinical trials. There are hospitals in this country who do not do a good job of diagnosing autism, trained behavioral biologists who don't know how to administer these tests properly. I know this because we can see healthcare sites that are diagnosing autism and others that are not trying to use the same standardized test. Take this into developing societies and it becomes very challenging to really know who has autism and who doesn't. That's not where it would go.

There's a lot of data now emerging on the complexity and the diversity of the microbiome is based on geography, largely based on diet, obviously legacy effects as well. There's a lot of evidence correlating diversity of microbial species to health. A lot of this does come from broad surveys of large populations of people, in let's say, the United States looking at 10,000 microbiomes, looking at their diet and lifestyle and correlating that metadata with microbiome profiles. I think there is very little debate that more diverse microbiomes correlate with health better than less diverse microbiomes. How do you get that? One or the other. Why are some people with less diverse microbiomes and more essentially with disease? We can speculate. Modern lifestyle, hygiene, drugs, preservatives in food, what have you. It's a difficult question to answer experimentally because you can get across a timescale of years and decades across large heterogenous populations.

The correlation to autism will be a very loose one, but there are fairly good correlations between microbiome diversity and complexity, and let's say, allergic diseases. This is really where the hygiene hypothesis originated in 1989. Essentially, the more people are exposed to microbes the more protected they are from allergies. There are now a number of different ways people have demonstrated that to be the case. Not just experimentally in mice, but in large human studies, mostly observational studies. I think you can correlate that question of healthy microbiomes, meaning more diverse microbiomes, to particular disease states that are easier diagnosed and probably more prevalent as well which makes it easier to study. Thirty percent of the world now has allergies of some sort.

ZIERLER: To go back to the laboratory experiments and just how difficult this research is, I can understand how you can study the gut of a mouse, but when you're looking at behavioral effects, a neurological disorder, how do you determine whether or not a mouse has autism? What does that even look like?

MAZMANIAN: That's a good question. I don't think mice get autism at all. I think autism is uniquely a human trait. If you look at what really goes into autism it's not just the behaviors. It's really sort of what is triggering those behaviors which are very hard to model in mice. Oftentimes there are sensory issues, attention issues, anxiety that is leading the child to not vocalize. And neuromuscular issues, physiological issues that are leading the child not to vocalize or want to vocalize, not to socialize or want to socialize. I'll take it one step back. The core behaviors by which autism is diagnosed are verbal and nonverbal social interactions. We communicate a lot without using words as well. Just reading a person's facial expression or knowing if they're in a bad mood somehow without them ever saying a word. That's difficult for children on the spectrum. Also it's called repetitive behaviors, but it's really more soothing or comforting behaviors. They're doing things over and over because they're familiar. That's comforting. That's how it's diagnosed. What's underlying that are sensory issues, anxiety, and issues with attention. How do you model all that in a mouse?

You can model whether the mouse is vocalizing or not, but you can't model why it stopped vocalizing. In the genetic models if there was a trigger that you can transport from a human DNA sequencing dataset to a mouse, meaning that I make a mutation in a mouse that I find in humans and the mouse behaves exactly like the human, then I'm more confident that the mouse is modeling human traits. But again, those don't really exist. The best mouse models a small fraction of what you would see in terms of diversity of behaviors in humans. There is no mouse model of autism and that's probably true for many other human disorders. What I do think you can do is model certain specific behaviors that are either hallmark behaviors of autism or otherwise associated with autism.

For example, anxiety is fairly robust to model in mice. If I truly believe a mouse is anxious it's still hard for me to know that. Really unravelling the underlying biology would ask: Is that mouse anxious because of the same reasons that a human being would be anxious? Were there different pathways leading to this behavior? That also posed a challenge to drug development and to starting biotech companies. To what degree can you even rely on the mouse data? I'm of the opinion now as my thoughts have evolved over the past several years that if you can do research in humans for autism you should. We should be very cautious in our interpretation of mouse research as it pertains to autism. Maybe not as it pertains to social behaviors, let's say. But the social behaviors associated with autism, whether or not that's telling us something about autism or just sociability…I think that's very much debatable.

ZIERLER: I can't help but ask if you've ever engaged with David Anderson about the emotional behaviors of animals and gauging them in the laboratory environment.

MAZMANIAN: David and I are friends and colleagues. We've had casual conversations. The animal models…once you accept their limitations and also divorce yourself and try to model human behaviors in mice, you can study a mouse. There's a lot of biology going on there. In terms of the circuits that would be shared between mice and humans, I think David would be able to comment much more on to what degree there is evolutionary conservation between the two systems. I would venture to guess that a lot of what's been discovered and really even believed about mice we are still gathering evidence to what degree the underlying neural circuits are at play in humans. It's not easy and oftentimes there is no experiment that you could do in humans to ever look at the nuances of brain activity, neural activity…they wouldn't be able to do fairly easily in mice. You can't even do the experiment in mice. Even postmortem you can't do the experiment. Maybe another way of saying this is, mice may give us go-no-go answers to whether or not the microbiome or you name it—a drug—has relevance to a certain emotional behavior. I think ultimately that's a springboard to human studies, if possible.

ZIERLER: This political minefield surrounding autism…from a scientific perspective the people who say, "There's nothing wrong with me. I can function perfectly well in society"…is it even from a scientific basis, is the metaphor of a spectrum, is that even the right way to think about it? In other words, for the people that you're motivated to help, the severe autism where it's obviously some kind of a condition that they would like relief for. Is it even useful to think about autism as a single issue where there is a spectrum of intensity? Why not think about it as the people who suffer so badly their lives are impacted negatively? Why don't we think of that as them having a different kind of illness than the people who identify as autistic and say, "I'm not ill. There's nothing wrong with me." How is the best way to think about these things?

MAZMANIAN: It's a very good point. I'm on the sidelines about this and I'm a mouse researcher. There are many people who deal with this more directly. They are trying to address the question that you've asked. The fact that we arrived from…even before autism…it's not coming to me. There was a term that people used that is widely believed to have…was hung by the same group of individuals. The fact that we've transitioned from autism to autism spectrum disorder was to address this particular point is that it is such a heterogenous disease. You find people where you can spend a day with them and never know that they have a diagnosis because the diagnosis is invented by somebody— [laugh]

ZIERLER: Right.

MAZMANIAN: —who put the questionnaire together. And other people that within two seconds you know something is going on. To what degree you can parse that out—this is an active debate in the field. It's driven a lot by the patient population itself, but also scientists and psychologists and parents who are very involved in shaping how autism is diagnosed, if not treated. There is a movement to re-characterize and rename the disorder. This is not a trivial thing to do. It's very difficult to get buy-in from everybody. You bring these international panels together to try and motivate people to follow the recommendations because you have enough people from enough groups that are of the same mind of how you label a person and where they are on the disorder. The active dialogue now is to rename—and I alluded to this earlier in the conversation—those individuals who have very severe symptoms as profoundly autistic. That is the actual term people are now suggesting they use to differentiate people who do need care from those who do not need care.

I have no idea where this discussion is going to go. It's always changing and it's a mess if I may just unload somewhat. I worry that the mess is impeding research. It's making lives difficult for parents and the affected individuals in terms of just societal views. In a way that I don't see happening—we've worked on a number of different disease areas. We work on mice, but we always are talking to physicians and scientists who are studying the human disorder. There is a lack of uniform acceptance on how to address autism and a lot of contentious views that are not shared in other patient populations. There is something different—and maybe for the reasons I mentioned a few minutes ago— about studying autism, and it's not doing many people a service in getting caught up in labels.

ZIERLER: Given that your thinking about the source of autism has changed over the years; it's such a different dichotomy: Professor Mazmanian at Caltech…it's fine for your thinking to change. You're just going to follow the data wherever it goes and you're going to adjust your research accordingly. But Dr. Mazmanian as biotech founder where there is a fiduciary responsibility—what does that mean for Axial if the foundation of the company is based on all of this excitement that we're really going to focus on the gut as the source of autism? What does that mean from a financial, from a company, from a commercial perspective when the founding researcher…his own thinking changes over the years?

MAZMANIAN: I don't think there is a material effect on the company because to be perfectly honest I'll say what I just said to you. I'll say it to anybody. I'm not sure investors will appreciate or really even want to hear this. Not to be discouraging. I think they're more interested in really very basic connections. Is there a target? Can I drug it? Can I make money off it? What it all means is secondary to what can actually be done. [laugh] In terms of what can be done I don't think the change in thinking changes the company at all. The change in thinking has gone from the microbiome is causal to the microbiome is one of multiple contributing factors, but it is a required contributing factor. Meaning that if I can modify the microbiome and restore a healthy state I can still improve outcomes. It's not the sole driver; it's a required driver of disease outcomes.

Our experiments after this shift in thinking in mice are bearing this out and the results in humans—regardless of what I think—they are their own results. They still look quite promising. In terms of the change of thought it may change maybe future approaches, but it's not going to change the current approach to autism. I don't think it diminishes from that approach as well. If we got to the point—and this may happen—where we are now convinced the microbiome has no relevance to human autism; obviously that's a game-changer in terms of how you would approach this from a drug development standpoint or what it means to a biotech company. But we're not there yet. I think it does have a role.

I just think…again maybe a way to rephrase this, is I don't think microbiome changes are specific to autism. I don't think there are specific changes. I think there are just healthier and less healthy microbiome states. If you're predisposed to autism and your microbiome is less healthy you're going to get autism. You're not going to get Parkinson's disease. But if you're predisposed to diabetes and your microbiome is less healthy you're going to get diabetes and not autism. I think the unhealthiness of the microbiome is addressable across multiple disease states.

ZIERLER: So that's to say that the current state of play is that if the gut microbiome contributes to the disease it's still feasible to create drugs that will be therapeutic, that will actually have a clinical value. That's sort of where Axial is right now. That's what hasn't changed since the beginning.

MAZMANIAN: The approach hasn't changed because whether or not it's the cause or a contributor, the molecular discoveries are what they are. Their effect size are what—and do they work in isolation is what has changed. Going back to the mouse data. What we found in mice on a molecular and cellular level very nicely translated to humans. Once we looked for the same changes that we discovered in mice which no one had ever looked at before we saw those changes. The target, if you will, is still there. The question that Axial will most likely answer in the next year or so is, If you now drug that target do you get robust and validated improvements in features or symptoms that are relevant to autism? The data look encouraging at this point, but we don't have all the statistics, and we don't have—very importantly—the control placebo group which we're enrolling now. You can imagine for, let's say, a cancer where you can measure a tumor. That's pretty objective. The tumor is getting smaller. It's getting smaller because it is, not because I'm thinking it's getting smaller. Any psychiatric disorder…there are huge placebo effects.

ZIERLER: Yes.

MAZMANIAN: In fact, people study the placebo effect because it's actually real. [laugh]

ZIERLER: Right. [laugh]

MAZMANIAN: People who think they're on a placebo; they actually are getting better, but it's not because of the drug. It's because they are changing their neurochemistry somehow which is exciting to study on a biological level. Until we have that control placebo group for a psychiatric disorder you don't really know if the effect size that you're observing in the patient population is relevant to overcome the placebo effect. Within the next year Axial will have that answer. If it doesn't turn out that the microbiome is contributing in a meaningful way I think that…and I'm hopeful that Axial's approach and the drug help people. But I think the long term benefit of that is that more and more people will now get into autism research. It's unlikely we've stumbled upon the answer; it's probably not the answer. With more successes in a particular space more people will try their approaches, their angles, their hypotheses and they won't be dissuaded by the barrier to entry in a very challenging disease area.

ZIERLER: The timeline looking at a year thinking about the placebo effect. Is this about where you thought you would be at the founding of Axial in terms of just expectations, how long things take?

MAZMANIAN: I can't say I ever thought about the timelines or had expectations or objectives for those timelines. What I can say retrospectively is if you look at other companies and how long they've taken even in more traditional areas where there is a lot more groundwork that was done previous to whatever that company was trying to do. For example, there's already therapeutics and they're just trying to develop a new therapeutic. Looking back in time for a small biotech to go from initial fundraising to clinical results which happened with us in three years—the first clinical data came in within three years—it's quite astounding actually. To be in this third clinical trial…it hasn't even been six years yet since the founding of the company. That's unusual and unusually fast as well. I can take zero credit for this. This is the team at Axial; they do a great job because they know how to execute in getting drugs into people.

I'll give them some more credit as well and this is not self-serving. Maybe this is selected by the employees who choose to work at Axial. They believe in the approach. Something in their mind makes them feel like this is a bet they want to take with their careers. I interact with the team at Axial quite often. They're just a very, very motivated, enthusiastic group of people. Maybe it's self-selecting. They could've chosen to work…maybe those who thought this was crazy never took the job at Axial. But it takes that as well. Maybe in some ways just this determination, stubbornness, whatever it may take—belief, maybe sometimes unfounded—to be able to push through the uncertainty of going into a new area.

ZIERLER: Going back to 2015 and the chronology with the founding of Axial. At the same time you were named an HMRI Investigator, the Heritage Medical Research Institute. What is that institute and why was that valuable for you to take on that affiliation?

MAZMANIAN: This is Richard Merkin's gift to Caltech. He set up the HMRI program; the Heritage Medical Research Institute. For me, personally—and I've had a relationship with Dr. Merkin and in terms of full disclosure he's an investor in Axial. His fund is an investor in Axial, just to disclose that as well. He's also a friend and he supported our research prior to the HMRI program. In fact this is a good connection. Probably 2009 or 2010 when Paul [Patterson] and I had first started generating this data…I might have told you this story. We had probably put in a couple of grant applications that hadn't been funded and then I was chatting with the provost about this project. Paul and I chatted with the provost, Ed Stolper at the time, about finding a new way to fund this crazy idea that the gut and the microbiome impact autism mouse models. Less than a couple months later Ed told us that he had found a donor who decided to give us $400,004—and I don't know how they came up with that number, but that was the number. That was Richard Merkin. He just took a chance on this because Ed and he were having a conversation and Ed brought them—it's the provost's job, right? He was a believer way back when. That resulted in that initial funding for this project which led to discoveries which led to the founding of a company that he later invested in.

The HMRI program itself has nothing to do with the investments of the company, but it's a program as you know that funds individual scholars here at Caltech. What it meant to me—and I think most would share this view—is the fact that the funds are unrestricted, meaning that they're essentially funding the lab and not funding projects, which is very different than most funding agencies. How most funding mechanisms work is that you pitch an idea. Of course the investigator is evaluated as part of the grant application, the track record, and what have you. Really it's the idea that's getting funded. In some cases, it's like a very mature research project, it's like almost fully validated before it gets funded. Whereas unrestricted funds are here as an allocation of funds; it's not assigned to any one particular project—you use it as you will.

That allows us to take risks. That allows us to really rapidly move into new areas. It tests hypotheses very quickly without the ramp up of trying to get funding for a project which could take a year or more even when it's successful. That's allowed us to really both diversify what we do and move rapidly in areas that may seem out of the box to traditional funding agencies, but something that we wanted to do. Very little reporting which is great and also—and I'm not sure this is embedded into the program, but I think it's probably a product of that. It increases our interaction with Dr. Merkin to the point where, through that dialogue will come things like investment in the company or maybe additional sources of money or connections to prominent scientists that we wouldn't have had without that relationship. There has been for me and I think many others in the program value in just building our relationship with Dr. Merkin.

ZIERLER: Were you thinking about Parkinson's at the same time as autism? Were you developing these ideas basically in parallel?

MAZMANIAN: The chronology is autism research came first because of the interactions with Paul [Patterson]. As we got more bold in our ability to ask questions in the gut-brain research space we went to Parkinson's because of the very well-known gastrointestinal manifestations in Parkinson's, and upwards of 80% of Parkinson's patients have constipation. That told me maybe the microbiome was involved. The Parkinson's work was started four or five years later. As I just mentioned they converged when the company was coming together just because our initial discovery of the link to the microbiome in Parkinson's mouse models came as the company was coming together. It just seemed like a natural home for that technology as well, and I certainly wasn't going to go start up another company on Parkinson's. Not just because of the heavy lift of starting any company which is real, but why would I start two companies that are essentially competing with each other in the gut-brain space? It's better to have more shots on goal in one company. That's how the connections to the company came was just because the timing just worked out.

If let's say the Parkinson's work hadn't matured until a year after the founding of Axial I'm not sure where that technology would be in terms of its ability to translate it into humans. I'm not sure where that would be today. I can tell you Axial did run a trial in Parkinson's as well. It was aborted because of COVID—elderly people stopped coming into the clinic, understandably—but they are restarting that program now as well. I think bundling the two into one company made a lot of sense and it continues to make sense.

ZIERLER: Is Parkinson's in some way a more satisfactory disease to focus on because it's more well-defined? It doesn't have these spectrum issues that autism has.

MAZMANIAN: There are surely pros and cons. Comparing those two disease areas in terms of what one may be able to reasonably expect from a therapeutic. We talked about all the issues in autism. In Parkinson's, parts of the etiology, the cause, are pretty well known. About ten percent is genetic; the mutations are well characterized. They run in families and the diagnosis runs in families. There's a very strong correlation genetically. Eighty-five to ninety percent is idiopathic meaning we still don't know what the cause is.

What is fairly common may not be the actual trigger, but the well-accepted pathology that leads to Parkinson's disease is that there is aggregation of a neuronal protein called alpha-synuclein. When alpha-synuclein changes its normal structure and starts to clump up inside cells and really damage and cause dysfunction in cells—neurons specifically—then that leads to Parkinson's. That actual commonality of a particular path of physiology that is quite widespread—it's not going to be 100%, but I'm going to venture well over half, if not most the vast majority of people with Parkinson's have the alpha-synuclein aggregation…at least gives you a target. Many of the therapeutics now are the more monotherapeutics aimed at disaggregating these alpha-synuclein aggregates. We have good targets to shoot at. There's many other types. In fact, the mainline therapy is just supplementing the type of neurons that are lost.

From that aspect it's much more feasible to think about Parkinson's as a disorder where the microbiome may be involved because also the gastrointestinal symptoms that are associated and less ambiguity in diagnosis. The cons are by the time a person—maybe this isn't a con, but this is just like the constraints of any therapeutic. By the time a person is diagnosed with Parkinson's it's been estimated that they've lost over 60% of a particular type of neuron in a specific structure of the brain. It's predicted that they've lost 60% of their dopaminergic neurons in their substantia nigra. Those neurons are likely not coming back with any therapy unless you're thinking about stem cell transplantations. Until those neurons are gone people don't have symptoms to be diagnosed. Now what you're working in is a window where you're unlikely to ever get a cure; you're unlikely to ever reverse symptoms. It's still very valuable to treat, but the goal is to slow down the progression and the holy grail is to stop the progression, but you're not going to reverse symptoms.

If we ever get to the day where we understand autism well enough to be able to treat it I believe those people will be able to learn new skills that they hadn't prior to that treatment. I think you can get benefits, I think you can get improvements in autism where you're unlikely to ever get improvements unless there is a way to regenerate neurons in Parkinson's. There are certainly attractive and less attractive features for both. I think if we went down a list of different diseases then we'd probably come up with the same pros and cons.

ZIERLER: What's the timeline for Parkinson's? You mentioned a year to get to the placebo studies for autism. What is the timeline look like for therapies on Parkinson's?

MAZMANIAN: It will probably be another year before we can start the clinical trials and that's going to require more fundraising as well. I don't know the answer to this, but the other aspect of why Parkinson's is more attractive than autism is what happens when there are already therapeutics for a disease. Not only does it de-risk investors and clinicians, but anyone else involved from pursuing that. What you also now have is a blueprint for what a clinical trial looks like because you've had a success in that clinical trial. You need to enroll a certain number of people, follow them for a certain amount of time, stop the drug, look at them for a certain amount of time afterwards to see if their symptoms rebound. The blueprint for how you conduct the trial exists for Parkinson's in ways it just simply doesn't exist for in autism because nothing has ever satisfied a therapeutic for autism.

I don't know what the answer is, but it's likely years. The major obstacle is enrollment. How do you get a sufficient number of people, 200-300 people, into a trial? There are real obstacles to that. It's not just that the patients exist, but do they want to be on an experimental trial? Even more fundamental—and this is TMI, but it's really interesting—clinicians that run clinical trials, they're running multiple clinical trials at the same time. They're just allocating. I have five patients with Parkinson's; I put one each on five different trials as opposed to bringing five people on one trial. It takes forever just to get the people and with a disorder of aging where you're unlikely to see acute or rapid changes for you to see any benefits you have to follow people over a period of time. Again, without having the answer I would venture to guess that at minimum these trials should last two or three years.

ZIERLER: Sarkis, given the sequencing that it was autism and then you had the idea—why not bring Parkinson's into the mix—does that open up a certain Pandora's box of saying why not look at all CNS disorders? In other words are there any CNS disorders that would not be amenable to the kind of approach that Axial has taken?

MAZMANIAN: Certainly, because there are many CNS disorders which haven't been studied well enough or haven't been linked to the microbiome. What is also emerging is that there are certain neuropsychiatric conditions which really appear to have a lot of close links to the microbiome and these are observational studies in humans. For example depression. Just in the past year or two a number of studies have come out which really showed very characteristic changes in the microbiome in people who have major depressive disorder and can rationalize how those changes in the microbiome are actually changing the neurochemistry in the brain. Now you have a testable hypothesis that originated from humans and didn't originate from mice, so I think that's a great place to start.

The other end of the argument which is true as well is that data is emerging suggesting that Alzheimer's is likely not highly linked to the microbiome. There is human data if you go and just do observational studies. Is the microbiome different in people with Alzheimer's versus mild cognitive impairment versus healthy individuals who are aged and match in other ways? There aren't huge changes in the microbiome. When we and others have looked in mouse models, we currently have a large NIH grant to study how the microbiome may contribute to Alzheimer's, the data is not as strong as the data we have in Parkinson's or even autism that the microbiome is involved. The data is not in any way as strong as the data we have in Parkinson's or even autism. Large whole cell changes in the microbiome are not moving disease endpoints that are linked to Alzheimer's meaning learning and memory tasks. Whether or not that's a feature of the mouse model microbiome I don't really know, but again we do have these…the first experiments as I described to you in autism showed us a connection. The first experiments in Parkinson's showed us a connection. Three years of experiments in Alzheimer's are not showing much in the way of connections. Others are finding the same thing. I think as the data comes out, people just start poking around in other areas thinking we are going to find more of these types of stories where certain disorders likely don't have at least an obvious microbiome component to that.

ZIERLER: Sarkis, some of your recent work on cancer—specifically colon cancer looking at probiotics and things like that—just in the way that you organize your research agenda is that a totally different area than looking at the neurological issues?

MAZMANIAN: That was a one-off paper because it's not the focus of our lab. There are a lot of people working on cancer in the microbiome, where there is tons of compelling data and tons of compelling data in humans which obviously has a much more powerful…there was a postdoc in our lab who was interested in making a link to inflammation associated cancers. This is Yun Kyung Lee who is now a professor in South Korea. What was previously known is that chronic inflammation is a risk factor for colon cancer. Inflammation changes the physiology of epithelial cells which then can cause mutations and lead to cancers. There are other mechanisms that can lead to colon cancers as well. Her hypothesis is if she were to block or suppress that inflammation with a probiotic could she prevent the development of a cancer in a mouse model that is immune driven to have colon cancer and the answer was yes.

As striking as that finding was, the fact that it worked conceptually—because the evidence existed that inhibiting inflammation is a way to prevent the development of cancer—it made a lot of sense as well based on the biology. It's not something we followed. We routinely will patent ideas that we either don't follow up on or nobody follows up on. I'm not even sure what the state of that technology is, meaning the IP.

ZIERLER: Bringing our conversation closer to the present both for your lab at Caltech and for Axial, what was the impact on COVID and all of the restrictions it created?

MAZMANIAN: I mentioned on the Axial side the clinical trial that was forced to be halted. We were no longer enrolling Parkinson's patients in clinical trials. Other than that, not much of a real day-to-day effect on Axial other than what we all went through. Social distancing and masks and what have you. From the research side nothing specific to our lab, but what all labs at Caltech and most places in the world had to do is for at least a brief period was to shut down. Obviously it's challenging when you have experiments that last in some cases over a year. In the middle of these to shut things down we had to make some very difficult decisions. There were entire experiments that were never completed because of COVID. Even though Caltech did a great job of allowing us to come in and finish experiments that we had a lot of investment in, experiments that were running for many months, but then needed to be analyzed in that acute phase between March and June of 2020.

The other effect, not specific to our lab, but when you're working with mice, it's not as if the day that you come in the mice are just ready for you to start experimenting on. There's a several-month ramp up to getting our mouse lines and the colonies bred where we have sufficient numbers of animals to do experiments. That ramp up in some projects took well over six months for them to even get to where they were when we shut down. I think everyone in biomedicine had to go through this, at least in academia.

Just to maybe take this point in a different direction. It didn't affect me at all in terms of my career. My main concern was about the trainees in the lab and how it stalled their careers from a time standpoint let alone the emotional, the psychological standpoint. Let alone reevaluating where they are in the world and did they want to continue in academia or in science all together, which a lot of people went through as well. The collateral damage from this focus on academia was really the careers of graduate students and postdocs. I think most people in our lab have done well in recovering, but it's not ideal when you are in a phase of your life where you're trying to produce something to get to the next phase of your life. It's an unenviable place to be when something as catastrophic as a pandemic happens.

ZIERLER: It's an important point to recognize that people who are secure in their careers…that's at least one thing they didn't have to worry about. That's a really important point to raise. You said that a lot of the trainees have recovered. They're doing well now?

MAZMANIAN: They're resilient and they've done what they needed to do to fill the gaps and pick up where they left off to the best of their abilities. Look, if it was just their careers are delayed one or two years that's not optimal by any means, but that's something we can all deal with or people can deal with.

ZIERLER: Sure.

MAZMANIAN: The psychological part—I'll get specific in a minute—I think that's harder to reconcile than just a time component. For example, it took me a while, just to be fair, to realize the emotional issues and the mental health issues—not so much in my lab—but other people I was talking to through the pandemic were going through. A lot of people were in small apartments isolated by themselves. In some cases if they're international students their families are thousands of miles of way. Whereas for me I was just super fortunate. I had my family around, we have a backyard, we could take a walk, we could go outdoors. I think we were just more protected from the emotional aspects of the pandemic than trainees were because they don't have the financial means to surround themselves with things that can get their mind off the pandemic. They were really in a vulnerable position in their careers and their lives. Beyond just the work there was the personal aspect that was and may even remain challenging for a lot of people.

ZIERLER: Sarkis, I'm curious just because you're fearless in jumping into new areas of research. Is long COVID something that you're interested in? Is there any connection between the microbiome and all of the neurological disorders that seem to be manifesting? Is there work to be done in that area for you?

MAZMANIAN: Not for me. In fact I made the decision early on in the pandemic to not do COVID research. There were a lot of funds available both internally and externally to pivot to COVID research. I made the decision not to do that because many talented people obviously got into COVID research and I wanted to maintain the focus of our lab. Once you commit…that's why we didn't do the colon cancer research. I think a lab has to be defined in some ways. You can't do everything. You don't have the capabilities or know how to do everything. You have to lean on your strengths and know where your strengths are and build on them. We didn't dive into COVID research. I don't know what long COVID is. There is no clinical definition for long COVID. As problematic as questionnaires are—they may be invalid, but there's a score at the end of the day. That doesn't even exist for long COVID. I don't think you can tell a person who has long COVID from a person that doesn't. Someone may develop some arbitrary metric, but if you can't define it I'm not sure you can study it.

ZIERLER: Sarkis, moving right to the present. What are you currently working on these days as you try to keep your head above water?

MAZMANIAN: We continue to be really excited about Parkinson's for the reasons that I had described where the effects in mice are very robust. The translation to humans appears obvious and at least worth exploring. A portion of our laboratory continues to work on the gut-brain connection. We're now expanding to other questions on the hypotheses looking at different cell types, whether they're bacterial or neuronal or other brain cells. We've started this work on Alzheimer's as I mentioned. We've pivoted a little bit away from…we've refocused what we're doing in autism again on the genome-environment interactions. But also to now what I think at least makes sense to me is to focus on more digestible aspects of autism. Not to think about…as I said you can't model the whole disease, but I'm very confident we can model the anxiety portion of autism. By understanding more about anxiety we may be able to help people who have autism and anxiety.

More recently…in fact, we just published a paper this morning. It's on the Caltech website. We've started looking at how the microbiome affects certain reward behaviors. This first paper was about how the microbiome impacts the ability, or alters the ability of the animal to over-consume palatable foods. It turns out that if you deplete a mouse's microbiome they eat more, but not regular food. They eat the same amount of chow. But if you give them sugar pills, anything sweet…doesn't even have to be sugar. It could be Splenda or other sweeteners that have no caloric value. They pretty dramatically over-consume palatable or desirable foods. A student in the lab who just graduated, a BMB student named James Ousey, showed that there are particular microbes—this is a function of a handful of microbes—that would suppress the animal's desire for hedonic feeding. I think this is super interesting because it gets into not just diet and let's say obesity or diabetes and other metabolic disorders which have been linked to the microbiome, but into behaviors and motivation for why an animal does certain things.

From that we've now layered upon a very new project where we don't even have data yet, but a new project on looking at the microbiome in substance abuse disorder. I think the biology may be shared between that and feeding because the mice that we were manipulating the microbiome—they changed their feeding behaviors and weren't doing it for nutritional value. They were doing it for pleasure. I think this gets into this question which we don't have an answer to. Does the microbiome somehow tap into the reward centers in our brain and change our motivation for desirable things or even things that are bad for us like drugs, but they provide some level of pleasure? That's the new frontier for our lab.

ZIERLER: Sarkis, now that we've worked right up to the present, to wrap up this excellent series of discussions I want to ask a few overall reflective questions. Then we can end looking to the future. I wonder just as a thought experiment if you had the opportunity to go and visit yourself back when you were in graduate school and tell you about the future, the kinds of things that you would pursue in your career, do you think during graduate school you would be surprised at how involved you've become in translational research and business? Or do you think from your family background, from your academic training, from your personality, that that seemed like a natural progression for you after your Ph.D.

MAZMANIAN: That's a really good question. Thanks for asking me and forcing me to think about this. [laugh]

ZIERLER: [laugh]

MAZMANIAN: It's great for me. I did not ever predict when I was in graduate school the trajectory of my career. I actually had no plans. I think I might have told you this, but I've never been goal-oriented in my career or in anything in life. Nor did I ever think, "Oh. It'd be nice to do this or I could do this." Or something. But going back in time, I'm not surprised because I think that all the things that made me adventuresome in science or drawn to company building or motivated to help people were always there. Elements of that were always there. It's part who I am. I don't know why, but that's who I am. I don't think graduate school Sarkis would've been surprised. I can ask this to my colleagues who were in graduate school with me whom I keep in touch with who probably knew me better than I knew myself at the time. I doubt they'd be surprised either, but I'm going to ask the question to them.

Research Symbiosis in Business and at Caltech

ZIERLER: Do you view the duality of your career as being symbiotic? Is there a mutual value in having a hand both in the fundamental research and the translational research? In other words, does your lab work at Caltech benefit your startup work and vice versa? Do you see it in those terms?

MAZMANIAN: Absolutely. Again, that was not the design for getting into a translational research or translating our basic science discovery, but it has been a huge benefit in both directions that you described. Certainly what we discover in mice as I've talked about over and over is it gives us insights or questions to ask in the clinic. As a basic scientist interacting with the clinicians and the human biologists really benefited me.

What I'm now learning from interacting with patient populations and reading more about the lives of people with specific disorders…I just spent some time in an office in clinic to just get to understand what autism is and not just read about it. I think you could read for a decade and never really know what autism is unless you spend time with individuals who both treat and have autism. It is important to our research. I think this is something we do and should be doing more of in the lab, is really tapping into and mining our access to human data and more importantly human experiences. Then determining ways of modeling that in mice. Then at least you're closer to the more relevant questions, the things that are going to affect quality of life and the things that matter most to the patient population. Then really leverage the power of experimental models because there's a lot that you cannot do in humans because it's unethical or not feasible or you just can never marshal the resources to do the human studies until you de-risk the concept in mice. Instead of asking questions in a vacuum in experimental models why not have the way you set up experiments inform you about something you know is real in people?

I think we do that, we do more and more of that, have students who will think along these lines or tell me, "Please introduce me to certain clinicians who can tell me more about the human disorder." Or they attend conferences which are medical conferences and not research conferences just to learn more about the human condition. It's called reverse translation. I think the more we do of that the more on target we will be with our experiments.

ZIERLER: Sarkis, a Caltech institutional question. Obviously decades ago you would never be able to do the kind of startup work as a Caltech professor that you're now able to do. The big historical narrative there of course is Caltech's embrace of tech transfer, of a culture of entrepreneurialism. But the narrative isn't static. Things always move forward. Based on your powers of extrapolation what's the next chapter in Caltech's history? If we had pure fundamental research for so long now we're in this moment here where there are so many Caltech professors, you among so many others who are doing such exciting and important work. What's the next generation? What do the next 10, 20, 30 years look like for Caltech in this space?

MAZMANIAN: My opinions will diverge from some, maybe not most faculty here at Caltech. First of all, the institution itself has moved more towards bringing on enabling capabilities to do translational research. As you mentioned company building was something that had existed at Caltech for many decades before I started here and of course it's grown and changed. The biotech industry or the tech industry has changed and it's sort of changed with it. I think we do a great job, honestly, in terms of tech transfer and how they really equip the faculty with the resources, the knowledge, the connections we need to build companies.

What Caltech has also done—and this gets back to Richard Merkin as one example—is to build a translational science institute here, Merkin Translational Research Institute. Which provides a different, but complimentary skillset to faculty in terms of entrepreneurialism, but also in really connecting with clinicians and being able to access clinical samples or move into clinical trials or just interact with clinicians in and of itself. I think that that is a good balance for a place like Caltech.

As much as I think about human disease, I'm not sure how supportive I'd be of a medical center associated with Caltech. I think there are huge benefits to a basic science campus where we're not subject to the constraints that universities have when they have launched medical centers or like four football teams and other things that sort of distract you from the focus of discovery. I love the fact that here there's more bandwidth for students, postdocs, and faculty just to ask bold and daring questions which just become institutionally more challenging when you have bureaucratic entities like a medical center to deal with. Nothing is impossible because you can do whatever you want on your own bench with your own time. I think the culture changes with having a medical center. So I like the basic science feel to Caltech. I think bringing in some capabilities on campus, and I think we're fairly good. I'm not sure we need much more though if it makes sense we should. But also to increase those connections to people off of campus in many different ways, not just medical or translational, but all the other sectors we talked about as well—fundraising, manufacturing, drug development, regulatory, what have you. Having those people accessible to the Caltech community, to me that's the ideal balance to essentially equipping the people on campus to take their basic ideas into the clinic and move from the clinic and design their own experiments.

ZIERLER: If you can compare the satisfactions in basic science and translational research… Of course in the lab when you make a discovery, when you learn something new about nature, that's the satisfaction. That's the eureka moment. For you and where you are on the translational side just as a sort of capsule in time do you feel like you have already helped people or is that not yet possible because of the regulatory framework, because of the state of the science, because of the clinical trials yet to come? How do you measure what it means to help people and what the timeline looks like for that?

MAZMANIAN: I clearly haven't helped people yet. I'm hopeful that some of the activities that haven't yet started will ultimately do that. I'll take a conservative definition of helping people. That is having an approved drug. Even though one can debate whether or not any certain approved drugs are really helping people. That's a different conversation. To me, that would be the bar. That would be criteria that I would like to satisfy for me to actually make the claim that I'm helping people or believe that I'm helping people. It is something I would love to do before my career is over.

Through that one—and I think important definition—is to have an approved drug that came from an idea that I had or more accurately the work that was done by the students and the postdocs in the lab. To take an idea and make it real through experiments and through data. There is a traditional path to drug development and we're squarely on that path. You have an idea, you test in the laboratory, it works out, and then all the other things we've talked about down the road until clinical trials. Very rarely does it happen where someone eats a plant and their cancer goes away. You just sort of do an end around this traditional path. I think for me to say that I've helped people…I think we're on that road, maybe on a number of roads to achieving that. But I would love that, to be honest.

ZIERLER: Finally, Sarkis, looking to the future because you have that clear benchmark on the translational side. What is that benchmark for the lab at Caltech? What are the frontiers for you that you haven't yet tackled?

MAZMANIAN: I think about this a lot and I've always sort of come to the same conclusion. I like to run the lab with the rule that we have no rules. Anything is possible. Within that constraint is just like my career. I don't know what the lab will be doing five years from now. I can't project from today. As we reach inflection points or opportunities to make decisions, I think we just go where our excitement and enthusiasm takes us and then figure out how to get there based on putting together the technical capabilities to answer those questions. Meaning that we're really driven by exciting questions and we'll make those in real time and then learn how to test those questions through collaborations, through equipment and knowledge. Then hopefully those will manifest into nice discoveries. But I cannot tell you that five years from now or even one year from now this is what the lab will be doing. I'm just hopeful we're going to stay at the cutting edge.

What also really makes me the happiest—maybe I can ask myself that question and answer it—is I can tell you I get no joy or satisfaction out of publishing a paper or getting a grant or certainly not an individual award. But when students achieve their career goals or postdocs get jobs, this is my reward. This is what basically my reward centers…it makes me happy. Really that the lab is a training environment. I know we talked a lot about the outcome of that training and what that may mean to hopefully society, but maybe to companies, maybe even to me. But really everything is a product of good training and having an environment where students can follow their interests and have the capabilities to do so. I think all the magic comes from that.

ZIERLER: Isn't that the Caltech way? Isn't that the message you got from your job talk and that made you decide this was the place for you?

MAZMANIAN: It is. Even other influences like Ray Owen who you may not know.

ZIERLER: Oh. I know of Ray Owen. Absolutely.

MAZMANIAN: Ray would say this to me. He had an office in the basement of Kerckhoff. I was in an adjoining building, Church. I'd try and seem him as often as I could; he was on campus a lot my first few years. He would say this all the time, "Isn't it just fun to let the students do what they want?"

ZIERLER: Yeah. [laugh]

MAZMANIAN: "And just encourage them and just to train them and teach them, but don't tell them what to do?" It really resonated with me not just because I felt like it was the right thing to do, but mainly because it aligned with who I was. I was learning that about me when Ray was saying that. Ray was here four decades before I was here. It is the Caltech way. I think many labs are run that way as well. You just sort of don't get in the way of talented, intelligent people. If you do you're just slowing them down.

ZIERLER: Great things will happen. Sarkis, this has been an incredible series of discussions. I'm so thankful for all of your time and insight. Thank you so much.

MAZMANIAN: Thank you, David.

[END]